redis.c 92.3 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
/*
 * Copyright (c) 2009-2010, Salvatore Sanfilippo <antirez at gmail dot com>
 * All rights reserved.
 *
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions are met:
 *
 *   * Redistributions of source code must retain the above copyright notice,
 *     this list of conditions and the following disclaimer.
 *   * Redistributions in binary form must reproduce the above copyright
 *     notice, this list of conditions and the following disclaimer in the
 *     documentation and/or other materials provided with the distribution.
 *   * Neither the name of Redis nor the names of its contributors may be used
 *     to endorse or promote products derived from this software without
 *     specific prior written permission.
 *
 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
 * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
 * ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
 * LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
 * POSSIBILITY OF SUCH DAMAGE.
 */

#include "redis.h"
31
#include "slowlog.h"
32
#include "bio.h"
33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49

#include <time.h>
#include <signal.h>
#include <sys/wait.h>
#include <errno.h>
#include <assert.h>
#include <ctype.h>
#include <stdarg.h>
#include <arpa/inet.h>
#include <sys/stat.h>
#include <fcntl.h>
#include <sys/time.h>
#include <sys/resource.h>
#include <sys/uio.h>
#include <limits.h>
#include <float.h>
#include <math.h>
50
#include <sys/resource.h>
51
#include <sys/utsname.h>
52 53 54 55 56

/* Our shared "common" objects */

struct sharedObjectsStruct shared;

57
/* Global vars that are actually used as constants. The following double
58 59 60 61 62 63 64
 * values are used for double on-disk serialization, and are initialized
 * at runtime to avoid strange compiler optimizations. */

double R_Zero, R_PosInf, R_NegInf, R_Nan;

/*================================= Globals ================================= */

65 66 67
/* Alternate stack for SIGSEGV/etc handlers */
char altstack[SIGSTKSZ];

68 69 70
/* Global vars */
struct redisServer server; /* server global state */
struct redisCommand *commandTable;
71

72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96
/* Our command table.
 *
 * Every entry is composed of the following fields:
 *
 * name: a string representing the command name.
 * function: pointer to the C function implementing the command.
 * arity: number of arguments, it is possible to use -N to say >= N
 * sflags: command flags as string. See below for a table of flags.
 * flags: flags as bitmask. Computed by Redis using the 'sflags' field.
 * get_keys_proc: an optional function to get key arguments from a command.
 *                This is only used when the following three fields are not
 *                enough to specify what arguments are keys.
 * first_key_index: first argument that is a key
 * last_key_index: last argument that is a key
 * key_step: step to get all the keys from first to last argument. For instance
 *           in MSET the step is two since arguments are key,val,key,val,...
 * microseconds: microseconds of total execution time for this command.
 * calls: total number of calls of this command.
 *
 * The flags, microseconds and calls fields are computed by Redis and should
 * always be set to zero.
 *
 * Command flags are expressed using strings where every character represents
 * a flag. Later the populateCommandTable() function will take care of
 * populating the real 'flags' field using this characters.
97 98 99 100 101 102 103 104
 *
 * This is the meaning of the flags:
 *
 * w: write command (may modify the key space).
 * r: read command  (will never modify the key space).
 * m: may increase memory usage once called. Don't allow if out of memory.
 * a: admin command, like SAVE or SHUTDOWN.
 * p: Pub/Sub related command.
105 106
 * f: force replication of this command, regarless of server.dirty.
 * s: command not allowed in scripts.
107
 * R: random command. Command is not deterministic, that is, the same command
108
 *    with the same arguments, with the same key space, may have different
109 110 111 112
 *    results. For instance SPOP and RANDOMKEY are two random commands.
 * S: Sort command output array if called from script, so that the output
 *    is deterministic.
 */
113
struct redisCommand redisCommandTable[] = {
114 115 116
    {"get",getCommand,2,"r",0,NULL,1,1,1,0,0},
    {"set",setCommand,3,"wm",0,noPreloadGetKeys,1,1,1,0,0},
    {"setnx",setnxCommand,3,"wm",0,noPreloadGetKeys,1,1,1,0,0},
117 118
    {"setex",setexCommand,4,"wm",0,noPreloadGetKeys,1,1,1,0,0},
    {"psetex",psetexCommand,4,"wm",0,noPreloadGetKeys,1,1,1,0,0},
119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137
    {"append",appendCommand,3,"wm",0,NULL,1,1,1,0,0},
    {"strlen",strlenCommand,2,"r",0,NULL,1,1,1,0,0},
    {"del",delCommand,-2,"w",0,noPreloadGetKeys,1,-1,1,0,0},
    {"exists",existsCommand,2,"r",0,NULL,1,1,1,0,0},
    {"setbit",setbitCommand,4,"wm",0,NULL,1,1,1,0,0},
    {"getbit",getbitCommand,3,"r",0,NULL,1,1,1,0,0},
    {"setrange",setrangeCommand,4,"wm",0,NULL,1,1,1,0,0},
    {"getrange",getrangeCommand,4,"r",0,NULL,1,1,1,0,0},
    {"substr",getrangeCommand,4,"r",0,NULL,1,1,1,0,0},
    {"incr",incrCommand,2,"wm",0,NULL,1,1,1,0,0},
    {"decr",decrCommand,2,"wm",0,NULL,1,1,1,0,0},
    {"mget",mgetCommand,-2,"r",0,NULL,1,-1,1,0,0},
    {"rpush",rpushCommand,-3,"wm",0,NULL,1,1,1,0,0},
    {"lpush",lpushCommand,-3,"wm",0,NULL,1,1,1,0,0},
    {"rpushx",rpushxCommand,3,"wm",0,NULL,1,1,1,0,0},
    {"lpushx",lpushxCommand,3,"wm",0,NULL,1,1,1,0,0},
    {"linsert",linsertCommand,5,"wm",0,NULL,1,1,1,0,0},
    {"rpop",rpopCommand,2,"w",0,NULL,1,1,1,0,0},
    {"lpop",lpopCommand,2,"w",0,NULL,1,1,1,0,0},
A
antirez 已提交
138 139 140
    {"brpop",brpopCommand,-3,"ws",0,NULL,1,1,1,0,0},
    {"brpoplpush",brpoplpushCommand,4,"wms",0,NULL,1,2,1,0,0},
    {"blpop",blpopCommand,-3,"ws",0,NULL,1,-2,1,0,0},
141 142 143 144 145 146 147 148 149 150 151 152
    {"llen",llenCommand,2,"r",0,NULL,1,1,1,0,0},
    {"lindex",lindexCommand,3,"r",0,NULL,1,1,1,0,0},
    {"lset",lsetCommand,4,"wm",0,NULL,1,1,1,0,0},
    {"lrange",lrangeCommand,4,"r",0,NULL,1,1,1,0,0},
    {"ltrim",ltrimCommand,4,"w",0,NULL,1,1,1,0,0},
    {"lrem",lremCommand,4,"w",0,NULL,1,1,1,0,0},
    {"rpoplpush",rpoplpushCommand,3,"wm",0,NULL,1,2,1,0,0},
    {"sadd",saddCommand,-3,"wm",0,NULL,1,1,1,0,0},
    {"srem",sremCommand,-3,"w",0,NULL,1,1,1,0,0},
    {"smove",smoveCommand,4,"w",0,NULL,1,2,1,0,0},
    {"sismember",sismemberCommand,3,"r",0,NULL,1,1,1,0,0},
    {"scard",scardCommand,2,"r",0,NULL,1,1,1,0,0},
153
    {"spop",spopCommand,2,"wRs",0,NULL,1,1,1,0,0},
154
    {"srandmember",srandmemberCommand,2,"rR",0,NULL,1,1,1,0,0},
155
    {"sinter",sinterCommand,-2,"rS",0,NULL,1,-1,1,0,0},
156
    {"sinterstore",sinterstoreCommand,-3,"wm",0,NULL,1,-1,1,0,0},
157
    {"sunion",sunionCommand,-2,"rS",0,NULL,1,-1,1,0,0},
158
    {"sunionstore",sunionstoreCommand,-3,"wm",0,NULL,1,-1,1,0,0},
159
    {"sdiff",sdiffCommand,-2,"rS",0,NULL,1,-1,1,0,0},
160
    {"sdiffstore",sdiffstoreCommand,-3,"wm",0,NULL,1,-1,1,0,0},
161
    {"smembers",sinterCommand,2,"rS",0,NULL,1,1,1,0,0},
162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183
    {"zadd",zaddCommand,-4,"wm",0,NULL,1,1,1,0,0},
    {"zincrby",zincrbyCommand,4,"wm",0,NULL,1,1,1,0,0},
    {"zrem",zremCommand,-3,"w",0,NULL,1,1,1,0,0},
    {"zremrangebyscore",zremrangebyscoreCommand,4,"w",0,NULL,1,1,1,0,0},
    {"zremrangebyrank",zremrangebyrankCommand,4,"w",0,NULL,1,1,1,0,0},
    {"zunionstore",zunionstoreCommand,-4,"wm",0,zunionInterGetKeys,0,0,0,0,0},
    {"zinterstore",zinterstoreCommand,-4,"wm",0,zunionInterGetKeys,0,0,0,0,0},
    {"zrange",zrangeCommand,-4,"r",0,NULL,1,1,1,0,0},
    {"zrangebyscore",zrangebyscoreCommand,-4,"r",0,NULL,1,1,1,0,0},
    {"zrevrangebyscore",zrevrangebyscoreCommand,-4,"r",0,NULL,1,1,1,0,0},
    {"zcount",zcountCommand,4,"r",0,NULL,1,1,1,0,0},
    {"zrevrange",zrevrangeCommand,-4,"r",0,NULL,1,1,1,0,0},
    {"zcard",zcardCommand,2,"r",0,NULL,1,1,1,0,0},
    {"zscore",zscoreCommand,3,"r",0,NULL,1,1,1,0,0},
    {"zrank",zrankCommand,3,"r",0,NULL,1,1,1,0,0},
    {"zrevrank",zrevrankCommand,3,"r",0,NULL,1,1,1,0,0},
    {"hset",hsetCommand,4,"wm",0,NULL,1,1,1,0,0},
    {"hsetnx",hsetnxCommand,4,"wm",0,NULL,1,1,1,0,0},
    {"hget",hgetCommand,3,"r",0,NULL,1,1,1,0,0},
    {"hmset",hmsetCommand,-4,"wm",0,NULL,1,1,1,0,0},
    {"hmget",hmgetCommand,-3,"r",0,NULL,1,1,1,0,0},
    {"hincrby",hincrbyCommand,4,"wm",0,NULL,1,1,1,0,0},
A
antirez 已提交
184
    {"hincrbyfloat",hincrbyfloatCommand,4,"wm",0,NULL,1,1,1,0,0},
185 186
    {"hdel",hdelCommand,-3,"w",0,NULL,1,1,1,0,0},
    {"hlen",hlenCommand,2,"r",0,NULL,1,1,1,0,0},
187 188
    {"hkeys",hkeysCommand,2,"rS",0,NULL,1,1,1,0,0},
    {"hvals",hvalsCommand,2,"rS",0,NULL,1,1,1,0,0},
189 190 191 192
    {"hgetall",hgetallCommand,2,"r",0,NULL,1,1,1,0,0},
    {"hexists",hexistsCommand,3,"r",0,NULL,1,1,1,0,0},
    {"incrby",incrbyCommand,3,"wm",0,NULL,1,1,1,0,0},
    {"decrby",decrbyCommand,3,"wm",0,NULL,1,1,1,0,0},
A
antirez 已提交
193
    {"incrbyfloat",incrbyfloatCommand,3,"wm",0,NULL,1,1,1,0,0},
194 195 196
    {"getset",getsetCommand,3,"wm",0,NULL,1,1,1,0,0},
    {"mset",msetCommand,-3,"wm",0,NULL,1,-1,2,0,0},
    {"msetnx",msetnxCommand,-3,"wm",0,NULL,1,-1,2,0,0},
197
    {"randomkey",randomkeyCommand,1,"rR",0,NULL,0,0,0,0,0},
198 199 200 201
    {"select",selectCommand,2,"r",0,NULL,0,0,0,0,0},
    {"move",moveCommand,3,"w",0,NULL,1,1,1,0,0},
    {"rename",renameCommand,3,"w",0,renameGetKeys,1,2,1,0,0},
    {"renamenx",renamenxCommand,3,"w",0,renameGetKeys,1,2,1,0,0},
202 203 204 205
    {"expire",expireCommand,3,"w",0,NULL,1,1,1,0,0},
    {"expireat",expireatCommand,3,"w",0,NULL,1,1,1,0,0},
    {"pexpire",pexpireCommand,3,"w",0,NULL,1,1,1,0,0},
    {"pexpireat",pexpireatCommand,3,"w",0,NULL,1,1,1,0,0},
206
    {"keys",keysCommand,2,"rS",0,NULL,0,0,0,0,0},
207
    {"dbsize",dbsizeCommand,1,"r",0,NULL,0,0,0,0,0},
A
antirez 已提交
208
    {"auth",authCommand,2,"rs",0,NULL,0,0,0,0,0},
209 210
    {"ping",pingCommand,1,"r",0,NULL,0,0,0,0,0},
    {"echo",echoCommand,2,"r",0,NULL,0,0,0,0,0},
A
antirez 已提交
211
    {"save",saveCommand,1,"ars",0,NULL,0,0,0,0,0},
212 213
    {"bgsave",bgsaveCommand,1,"ar",0,NULL,0,0,0,0,0},
    {"bgrewriteaof",bgrewriteaofCommand,1,"ar",0,NULL,0,0,0,0,0},
214
    {"shutdown",shutdownCommand,-1,"ar",0,NULL,0,0,0,0,0},
215 216
    {"lastsave",lastsaveCommand,1,"r",0,NULL,0,0,0,0,0},
    {"type",typeCommand,2,"r",0,NULL,1,1,1,0,0},
217
    {"multi",multiCommand,1,"rs",0,NULL,0,0,0,0,0},
218
    {"exec",execCommand,1,"s",0,NULL,0,0,0,0,0},
219 220
    {"discard",discardCommand,1,"rs",0,NULL,0,0,0,0,0},
    {"sync",syncCommand,1,"ars",0,NULL,0,0,0,0,0},
221 222
    {"flushdb",flushdbCommand,1,"w",0,NULL,0,0,0,0,0},
    {"flushall",flushallCommand,1,"w",0,NULL,0,0,0,0,0},
223
    {"sort",sortCommand,-2,"wmS",0,NULL,1,1,1,0,0},
224
    {"info",infoCommand,-1,"r",0,NULL,0,0,0,0,0},
225
    {"monitor",monitorCommand,1,"ars",0,NULL,0,0,0,0,0},
226 227
    {"ttl",ttlCommand,2,"r",0,NULL,1,1,1,0,0},
    {"pttl",pttlCommand,2,"r",0,NULL,1,1,1,0,0},
228
    {"persist",persistCommand,2,"w",0,NULL,1,1,1,0,0},
A
antirez 已提交
229
    {"slaveof",slaveofCommand,3,"as",0,NULL,0,0,0,0,0},
230
    {"debug",debugCommand,-2,"as",0,NULL,0,0,0,0,0},
231
    {"config",configCommand,-2,"ar",0,NULL,0,0,0,0,0},
232 233 234 235
    {"subscribe",subscribeCommand,-2,"rps",0,NULL,0,0,0,0,0},
    {"unsubscribe",unsubscribeCommand,-1,"rps",0,NULL,0,0,0,0,0},
    {"psubscribe",psubscribeCommand,-2,"rps",0,NULL,0,0,0,0,0},
    {"punsubscribe",punsubscribeCommand,-1,"rps",0,NULL,0,0,0,0,0},
236
    {"publish",publishCommand,3,"pf",0,NULL,0,0,0,0,0},
237 238
    {"watch",watchCommand,-2,"rs",0,noPreloadGetKeys,1,-1,1,0,0},
    {"unwatch",unwatchCommand,1,"rs",0,NULL,0,0,0,0,0},
239
    {"cluster",clusterCommand,-2,"ar",0,NULL,0,0,0,0,0},
240
    {"restore",restoreCommand,4,"awm",0,NULL,1,1,1,0,0},
241
    {"migrate",migrateCommand,6,"aw",0,NULL,0,0,0,0,0},
242
    {"asking",askingCommand,1,"r",0,NULL,0,0,0,0,0},
243 244
    {"dump",dumpCommand,2,"ar",0,NULL,1,1,1,0,0},
    {"object",objectCommand,-2,"r",0,NULL,2,2,2,0,0},
245
    {"client",clientCommand,-2,"ar",0,NULL,0,0,0,0,0},
246 247
    {"eval",evalCommand,-3,"s",0,zunionInterGetKeys,0,0,0,0,0},
    {"evalsha",evalShaCommand,-3,"s",0,zunionInterGetKeys,0,0,0,0,0},
248
    {"slowlog",slowlogCommand,-2,"r",0,NULL,0,0,0,0,0},
A
antirez 已提交
249 250
    {"script",scriptCommand,-2,"ras",0,NULL,0,0,0,0,0},
    {"time",timeCommand,1,"rR",0,NULL,0,0,0,0,0}
251 252 253 254
};

/*============================ Utility functions ============================ */

255 256 257
/* Low level logging. To use only for very big messages, otherwise
 * redisLog() is to prefer. */
void redisLogRaw(int level, const char *msg) {
J
Jonah H. Harris 已提交
258 259
    const int syslogLevelMap[] = { LOG_DEBUG, LOG_INFO, LOG_NOTICE, LOG_WARNING };
    const char *c = ".-*#";
260
    FILE *fp;
261
    char buf[64];
A
antirez 已提交
262
    int rawmode = (level & REDIS_LOG_RAW);
263

A
antirez 已提交
264
    level &= 0xff; /* clear flags */
265
    if (level < server.verbosity) return;
266 267 268 269

    fp = (server.logfile == NULL) ? stdout : fopen(server.logfile,"a");
    if (!fp) return;

A
antirez 已提交
270 271 272
    if (rawmode) {
        fprintf(fp,"%s",msg);
    } else {
273 274 275 276 277 278
        int off;
        struct timeval tv;

        gettimeofday(&tv,NULL);
        off = strftime(buf,sizeof(buf),"%d %b %H:%M:%S.",localtime(&tv.tv_sec));
        snprintf(buf+off,sizeof(buf)-off,"%03d",(int)tv.tv_usec/1000);
A
antirez 已提交
279 280
        fprintf(fp,"[%d] %s %c %s\n",(int)getpid(),buf,c[level],msg);
    }
J
Jonah H. Harris 已提交
281 282
    fflush(fp);

283
    if (server.logfile) fclose(fp);
J
Jonah H. Harris 已提交
284 285

    if (server.syslog_enabled) syslog(syslogLevelMap[level], "%s", msg);
286 287
}

288 289 290 291 292 293 294
/* Like redisLogRaw() but with printf-alike support. This is the funciton that
 * is used across the code. The raw version is only used in order to dump
 * the INFO output on crash. */
void redisLog(int level, const char *fmt, ...) {
    va_list ap;
    char msg[REDIS_MAX_LOGMSG_LEN];

A
antirez 已提交
295
    if ((level&0xff) < server.verbosity) return;
296 297 298 299 300 301 302 303

    va_start(ap, fmt);
    vsnprintf(msg, sizeof(msg), fmt, ap);
    va_end(ap);

    redisLogRaw(level,msg);
}

A
antirez 已提交
304 305 306 307 308 309 310 311 312 313 314 315 316 317
/* Log a fixed message without printf-alike capabilities, in a way that is
 * safe to call from a signal handler.
 *
 * We actually use this only for signals that are not fatal from the point
 * of view of Redis. Signals that are going to kill the server anyway and
 * where we need printf-alike features are served by redisLog(). */
void redisLogFromHandler(int level, const char *msg) {
    int fd;
    char buf[64];

    if ((level&0xff) < server.verbosity ||
        (server.logfile == NULL && server.daemonize)) return;
    fd = server.logfile ?
        open(server.logfile, O_APPEND|O_CREAT|O_WRONLY, 0644) :
A
antirez 已提交
318
        STDOUT_FILENO;
A
antirez 已提交
319 320
    if (fd == -1) return;
    ll2string(buf,sizeof(buf),getpid());
321 322 323
    if (write(fd,"[",1) == -1) goto err;
    if (write(fd,buf,strlen(buf)) == -1) goto err;
    if (write(fd," | signal handler] (",20) == -1) goto err;
A
antirez 已提交
324
    ll2string(buf,sizeof(buf),time(NULL));
325 326 327 328 329
    if (write(fd,buf,strlen(buf)) == -1) goto err;
    if (write(fd,") ",2) == -1) goto err;
    if (write(fd,msg,strlen(msg)) == -1) goto err;
    if (write(fd,"\n",1) == -1) goto err;
err:
A
antirez 已提交
330
    if (server.logfile) close(fd);
A
antirez 已提交
331 332
}

333 334 335 336 337 338 339 340 341 342 343
/* Redis generally does not try to recover from out of memory conditions
 * when allocating objects or strings, it is not clear if it will be possible
 * to report this condition to the client since the networking layer itself
 * is based on heap allocation for send buffers, so we simply abort.
 * At least the code will be simpler to read... */
void oom(const char *msg) {
    redisLog(REDIS_WARNING, "%s: Out of memory\n",msg);
    sleep(1);
    abort();
}

A
antirez 已提交
344 345 346 347 348 349 350 351 352 353 354
/* Return the UNIX time in microseconds */
long long ustime(void) {
    struct timeval tv;
    long long ust;

    gettimeofday(&tv, NULL);
    ust = ((long long)tv.tv_sec)*1000000;
    ust += tv.tv_usec;
    return ust;
}

355 356 357 358 359
/* Return the UNIX time in milliseconds */
long long mstime(void) {
    return ustime()/1000;
}

360 361 362 363 364 365 366 367 368 369 370 371
/* After an RDB dump or AOF rewrite we exit from children using _exit() instead of
 * exit(), because the latter may interact with the same file objects used by
 * the parent process. However if we are testing the coverage normal exit() is
 * used in order to obtain the right coverage information. */
void exitFromChild(int retcode) {
#ifdef COVERAGE_TEST
    exit(retcode);
#else
    _exit(retcode);
#endif
}

372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401
/*====================== Hash table type implementation  ==================== */

/* This is an hash table type that uses the SDS dynamic strings libary as
 * keys and radis objects as values (objects can hold SDS strings,
 * lists, sets). */

void dictVanillaFree(void *privdata, void *val)
{
    DICT_NOTUSED(privdata);
    zfree(val);
}

void dictListDestructor(void *privdata, void *val)
{
    DICT_NOTUSED(privdata);
    listRelease((list*)val);
}

int dictSdsKeyCompare(void *privdata, const void *key1,
        const void *key2)
{
    int l1,l2;
    DICT_NOTUSED(privdata);

    l1 = sdslen((sds)key1);
    l2 = sdslen((sds)key2);
    if (l1 != l2) return 0;
    return memcmp(key1, key2, l1) == 0;
}

402 403 404 405 406 407 408 409 410
/* A case insensitive version used for the command lookup table. */
int dictSdsKeyCaseCompare(void *privdata, const void *key1,
        const void *key2)
{
    DICT_NOTUSED(privdata);

    return strcasecmp(key1, key2) == 0;
}

411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441
void dictRedisObjectDestructor(void *privdata, void *val)
{
    DICT_NOTUSED(privdata);

    if (val == NULL) return; /* Values of swapped out keys as set to NULL */
    decrRefCount(val);
}

void dictSdsDestructor(void *privdata, void *val)
{
    DICT_NOTUSED(privdata);

    sdsfree(val);
}

int dictObjKeyCompare(void *privdata, const void *key1,
        const void *key2)
{
    const robj *o1 = key1, *o2 = key2;
    return dictSdsKeyCompare(privdata,o1->ptr,o2->ptr);
}

unsigned int dictObjHash(const void *key) {
    const robj *o = key;
    return dictGenHashFunction(o->ptr, sdslen((sds)o->ptr));
}

unsigned int dictSdsHash(const void *key) {
    return dictGenHashFunction((unsigned char*)key, sdslen((char*)key));
}

442 443 444 445
unsigned int dictSdsCaseHash(const void *key) {
    return dictGenCaseHashFunction((unsigned char*)key, sdslen((char*)key));
}

446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486
int dictEncObjKeyCompare(void *privdata, const void *key1,
        const void *key2)
{
    robj *o1 = (robj*) key1, *o2 = (robj*) key2;
    int cmp;

    if (o1->encoding == REDIS_ENCODING_INT &&
        o2->encoding == REDIS_ENCODING_INT)
            return o1->ptr == o2->ptr;

    o1 = getDecodedObject(o1);
    o2 = getDecodedObject(o2);
    cmp = dictSdsKeyCompare(privdata,o1->ptr,o2->ptr);
    decrRefCount(o1);
    decrRefCount(o2);
    return cmp;
}

unsigned int dictEncObjHash(const void *key) {
    robj *o = (robj*) key;

    if (o->encoding == REDIS_ENCODING_RAW) {
        return dictGenHashFunction(o->ptr, sdslen((sds)o->ptr));
    } else {
        if (o->encoding == REDIS_ENCODING_INT) {
            char buf[32];
            int len;

            len = ll2string(buf,32,(long)o->ptr);
            return dictGenHashFunction((unsigned char*)buf, len);
        } else {
            unsigned int hash;

            o = getDecodedObject(o);
            hash = dictGenHashFunction(o->ptr, sdslen((sds)o->ptr));
            decrRefCount(o);
            return hash;
        }
    }
}

487
/* Sets type hash table */
488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503
dictType setDictType = {
    dictEncObjHash,            /* hash function */
    NULL,                      /* key dup */
    NULL,                      /* val dup */
    dictEncObjKeyCompare,      /* key compare */
    dictRedisObjectDestructor, /* key destructor */
    NULL                       /* val destructor */
};

/* Sorted sets hash (note: a skiplist is used in addition to the hash table) */
dictType zsetDictType = {
    dictEncObjHash,            /* hash function */
    NULL,                      /* key dup */
    NULL,                      /* val dup */
    dictEncObjKeyCompare,      /* key compare */
    dictRedisObjectDestructor, /* key destructor */
504
    NULL                       /* val destructor */
505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526
};

/* Db->dict, keys are sds strings, vals are Redis objects. */
dictType dbDictType = {
    dictSdsHash,                /* hash function */
    NULL,                       /* key dup */
    NULL,                       /* val dup */
    dictSdsKeyCompare,          /* key compare */
    dictSdsDestructor,          /* key destructor */
    dictRedisObjectDestructor   /* val destructor */
};

/* Db->expires */
dictType keyptrDictType = {
    dictSdsHash,               /* hash function */
    NULL,                      /* key dup */
    NULL,                      /* val dup */
    dictSdsKeyCompare,         /* key compare */
    NULL,                      /* key destructor */
    NULL                       /* val destructor */
};

527 528 529 530 531 532 533 534 535 536
/* Command table. sds string -> command struct pointer. */
dictType commandTableDictType = {
    dictSdsCaseHash,           /* hash function */
    NULL,                      /* key dup */
    NULL,                      /* val dup */
    dictSdsKeyCaseCompare,     /* key compare */
    dictSdsDestructor,         /* key destructor */
    NULL                       /* val destructor */
};

537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558
/* Hash type hash table (note that small hashes are represented with zimpaps) */
dictType hashDictType = {
    dictEncObjHash,             /* hash function */
    NULL,                       /* key dup */
    NULL,                       /* val dup */
    dictEncObjKeyCompare,       /* key compare */
    dictRedisObjectDestructor,  /* key destructor */
    dictRedisObjectDestructor   /* val destructor */
};

/* Keylist hash table type has unencoded redis objects as keys and
 * lists as values. It's used for blocking operations (BLPOP) and to
 * map swapped keys to a list of clients waiting for this keys to be loaded. */
dictType keylistDictType = {
    dictObjHash,                /* hash function */
    NULL,                       /* key dup */
    NULL,                       /* val dup */
    dictObjKeyCompare,          /* key compare */
    dictRedisObjectDestructor,  /* key destructor */
    dictListDestructor          /* val destructor */
};

A
antirez 已提交
559 560 561 562 563 564 565 566 567 568 569
/* Cluster nodes hash table, mapping nodes addresses 1.2.3.4:6379 to
 * clusterNode structures. */
dictType clusterNodesDictType = {
    dictSdsHash,                /* hash function */
    NULL,                       /* key dup */
    NULL,                       /* val dup */
    dictSdsKeyCompare,          /* key compare */
    dictSdsDestructor,          /* key destructor */
    NULL                        /* val destructor */
};

570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613
int htNeedsResize(dict *dict) {
    long long size, used;

    size = dictSlots(dict);
    used = dictSize(dict);
    return (size && used && size > DICT_HT_INITIAL_SIZE &&
            (used*100/size < REDIS_HT_MINFILL));
}

/* If the percentage of used slots in the HT reaches REDIS_HT_MINFILL
 * we resize the hash table to save memory */
void tryResizeHashTables(void) {
    int j;

    for (j = 0; j < server.dbnum; j++) {
        if (htNeedsResize(server.db[j].dict))
            dictResize(server.db[j].dict);
        if (htNeedsResize(server.db[j].expires))
            dictResize(server.db[j].expires);
    }
}

/* Our hash table implementation performs rehashing incrementally while
 * we write/read from the hash table. Still if the server is idle, the hash
 * table will use two tables for a long time. So we try to use 1 millisecond
 * of CPU time at every serverCron() loop in order to rehash some key. */
void incrementallyRehash(void) {
    int j;

    for (j = 0; j < server.dbnum; j++) {
        if (dictIsRehashing(server.db[j].dict)) {
            dictRehashMilliseconds(server.db[j].dict,1);
            break; /* already used our millisecond for this loop... */
        }
    }
}

/* This function is called once a background process of some kind terminates,
 * as we want to avoid resizing the hash tables when there is a child in order
 * to play well with copy-on-write (otherwise when a resize happens lots of
 * memory pages are copied). The goal of this function is to update the ability
 * for dict.c to resize the hash tables accordingly to the fact we have o not
 * running childs. */
void updateDictResizePolicy(void) {
A
antirez 已提交
614
    if (server.rdb_child_pid == -1 && server.aof_child_pid == -1)
615 616 617 618 619 620 621
        dictEnableResize();
    else
        dictDisableResize();
}

/* ======================= Cron: called every 100 ms ======================== */

622 623 624 625 626 627 628 629 630 631 632 633 634 635 636
/* Try to expire a few timed out keys. The algorithm used is adaptive and
 * will use few CPU cycles if there are few expiring keys, otherwise
 * it will get more aggressive to avoid that too much memory is used by
 * keys that can be removed from the keyspace. */
void activeExpireCycle(void) {
    int j;

    for (j = 0; j < server.dbnum; j++) {
        int expired;
        redisDb *db = server.db+j;

        /* Continue to expire if at the end of the cycle more than 25%
         * of the keys were expired. */
        do {
            long num = dictSize(db->expires);
637
            long long now = mstime();
638 639 640 641 642 643

            expired = 0;
            if (num > REDIS_EXPIRELOOKUPS_PER_CRON)
                num = REDIS_EXPIRELOOKUPS_PER_CRON;
            while (num--) {
                dictEntry *de;
644
                long long t;
645 646

                if ((de = dictGetRandomKey(db->expires)) == NULL) break;
647
                t = dictGetSignedIntegerVal(de);
648
                if (now > t) {
649
                    sds key = dictGetKey(de);
650 651 652 653 654 655 656 657 658 659 660 661 662
                    robj *keyobj = createStringObject(key,sdslen(key));

                    propagateExpire(db,keyobj);
                    dbDelete(db,keyobj);
                    decrRefCount(keyobj);
                    expired++;
                    server.stat_expiredkeys++;
                }
            }
        } while (expired > REDIS_EXPIRELOOKUPS_PER_CRON/4);
    }
}

663
void updateLRUClock(void) {
664
    server.lruclock = (server.unixtime/REDIS_LRU_CLOCK_RESOLUTION) &
665 666
                                                REDIS_LRU_CLOCK_MAX;
}
667

668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692

/* Add a sample to the operations per second array of samples. */
void trackOperationsPerSecond(void) {
    long long t = mstime() - server.ops_sec_last_sample_time;
    long long ops = server.stat_numcommands - server.ops_sec_last_sample_ops;
    long long ops_sec;

    ops_sec = t > 0 ? (ops*1000/t) : 0;

    server.ops_sec_samples[server.ops_sec_idx] = ops_sec;
    server.ops_sec_idx = (server.ops_sec_idx+1) % REDIS_OPS_SEC_SAMPLES;
    server.ops_sec_last_sample_time = mstime();
    server.ops_sec_last_sample_ops = server.stat_numcommands;
}

/* Return the mean of all the samples. */
long long getOperationsPerSecond(void) {
    int j;
    long long sum = 0;

    for (j = 0; j < REDIS_OPS_SEC_SAMPLES; j++)
        sum += server.ops_sec_samples[j];
    return sum / REDIS_OPS_SEC_SAMPLES;
}

A
antirez 已提交
693 694
/* Check for timeouts. Returns non-zero if the client was terminated */
int clientsCronHandleTimeout(redisClient *c) {
695
    time_t now = server.unixtime;
696 697 698 699 700 701 702 703 704 705 706

    if (server.maxidletime &&
        !(c->flags & REDIS_SLAVE) &&    /* no timeout for slaves */
        !(c->flags & REDIS_MASTER) &&   /* no timeout for masters */
        !(c->flags & REDIS_BLOCKED) &&  /* no timeout for BLPOP */
        dictSize(c->pubsub_channels) == 0 && /* no timeout for pubsub */
        listLength(c->pubsub_patterns) == 0 &&
        (now - c->lastinteraction > server.maxidletime))
    {
        redisLog(REDIS_VERBOSE,"Closing idle client");
        freeClient(c);
A
antirez 已提交
707
        return 1;
708 709 710 711 712 713
    } else if (c->flags & REDIS_BLOCKED) {
        if (c->bpop.timeout != 0 && c->bpop.timeout < now) {
            addReply(c,shared.nullmultibulk);
            unblockClientWaitingData(c);
        }
    }
A
antirez 已提交
714
    return 0;
715 716
}

717
/* The client query buffer is an sds.c string that can end with a lot of
A
antirez 已提交
718 719 720 721
 * free space not used, this function reclaims space if needed.
 *
 * The funciton always returns 0 as it never terminates the client. */
int clientsCronResizeQueryBuffer(redisClient *c) {
722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739
    size_t querybuf_size = sdsAllocSize(c->querybuf);
    time_t idletime = server.unixtime - c->lastinteraction;

    /* There are two conditions to resize the query buffer:
     * 1) Query buffer is > BIG_ARG and too big for latest peak.
     * 2) Client is inactive and the buffer is bigger than 1k. */
    if (((querybuf_size > REDIS_MBULK_BIG_ARG) &&
         (querybuf_size/(c->querybuf_peak+1)) > 2) ||
         (querybuf_size > 1024 && idletime > 2))
    {
        /* Only resize the query buffer if it is actually wasting space. */
        if (sdsavail(c->querybuf) > 1024) {
            c->querybuf = sdsRemoveFreeSpace(c->querybuf);
        }
    }
    /* Reset the peak again to capture the peak memory usage in the next
     * cycle. */
    c->querybuf_peak = 0;
A
antirez 已提交
740
    return 0;
741 742
}

743 744 745 746 747 748
void clientsCron(void) {
    /* Make sure to process at least 1/100 of clients per call.
     * Since this function is called 10 times per second we are sure that
     * in the worst case we process all the clients in 10 seconds.
     * In normal conditions (a reasonable number of clients) we process
     * all the clients in a shorter time. */
749 750
    int numclients = listLength(server.clients);
    int iterations = numclients/100;
751

752 753
    if (iterations < 50)
        iterations = (numclients < 50) ? numclients : 50;
754 755 756 757 758 759 760 761 762 763
    while(listLength(server.clients) && iterations--) {
        redisClient *c;
        listNode *head;

        /* Rotate the list, take the current head, process.
         * This way if the client must be removed from the list it's the
         * first element and we don't incur into O(N) computation. */
        listRotate(server.clients);
        head = listFirst(server.clients);
        c = listNodeValue(head);
A
antirez 已提交
764 765 766 767 768
        /* The following functions do different service checks on the client.
         * The protocol is that they return non-zero if the client was
         * terminated. */
        if (clientsCronHandleTimeout(c)) continue;
        if (clientsCronResizeQueryBuffer(c)) continue;
769 770 771
    }
}

772
int serverCron(struct aeEventLoop *eventLoop, long long id, void *clientData) {
773
    int j, loops = server.cronloops;
774 775 776 777
    REDIS_NOTUSED(eventLoop);
    REDIS_NOTUSED(id);
    REDIS_NOTUSED(clientData);

A
antirez 已提交
778 779 780 781
    /* Software watchdog: deliver the SIGALRM that will reach the signal
     * handler if we don't return here fast enough. */
    if (server.watchdog_period) watchdogScheduleSignal(server.watchdog_period);

782 783 784 785 786
    /* We take a cached value of the unix time in the global state because
     * with virtual memory and aging there is to store the current time
     * in objects at every object access, and accuracy is not needed.
     * To access a global var is faster than calling time(NULL) */
    server.unixtime = time(NULL);
A
antirez 已提交
787

788 789
    trackOperationsPerSecond();

790
    /* We have just 22 bits per object for LRU information.
791 792
     * So we use an (eventually wrapping) LRU clock with 10 seconds resolution.
     * 2^22 bits with 10 seconds resoluton is more or less 1.5 years.
793
     *
794
     * Note that even if this will wrap after 1.5 years it's not a problem,
795
     * everything will still work but just some object will appear younger
796 797 798 799 800
     * to Redis. But for this to happen a given object should never be touched
     * for 1.5 years.
     *
     * Note that you can change the resolution altering the
     * REDIS_LRU_CLOCK_RESOLUTION define.
801
     */
802
    updateLRUClock();
803

804 805 806 807
    /* Record the max memory used since the server was started. */
    if (zmalloc_used_memory() > server.stat_peak_memory)
        server.stat_peak_memory = zmalloc_used_memory();

808 809 810
    /* We received a SIGTERM, shutting down here in a safe way, as it is
     * not ok doing so inside the signal handler. */
    if (server.shutdown_asap) {
811
        if (prepareForShutdown(0) == REDIS_OK) exit(0);
812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833
        redisLog(REDIS_WARNING,"SIGTERM received but errors trying to shut down the server, check the logs for more information");
    }

    /* Show some info about non-empty databases */
    for (j = 0; j < server.dbnum; j++) {
        long long size, used, vkeys;

        size = dictSlots(server.db[j].dict);
        used = dictSize(server.db[j].dict);
        vkeys = dictSize(server.db[j].expires);
        if (!(loops % 50) && (used || vkeys)) {
            redisLog(REDIS_VERBOSE,"DB %d: %lld keys (%lld volatile) in %lld slots HT.",j,used,vkeys,size);
            /* dictPrintStats(server.dict); */
        }
    }

    /* We don't want to resize the hash tables while a bacground saving
     * is in progress: the saving child is created using fork() that is
     * implemented with a copy-on-write semantic in most modern systems, so
     * if we resize the HT while there is the saving child at work actually
     * a lot of memory movements in the parent will cause a lot of pages
     * copied. */
A
antirez 已提交
834
    if (server.rdb_child_pid == -1 && server.aof_child_pid == -1) {
835 836 837 838 839 840 841 842 843
        if (!(loops % 10)) tryResizeHashTables();
        if (server.activerehashing) incrementallyRehash();
    }

    /* Show information about connected clients */
    if (!(loops % 50)) {
        redisLog(REDIS_VERBOSE,"%d clients connected (%d slaves), %zu bytes in use",
            listLength(server.clients)-listLength(server.slaves),
            listLength(server.slaves),
844
            zmalloc_used_memory());
845 846
    }

847 848
    /* We need to do a few operations on clients asynchronously. */
    clientsCron();
849

850 851
    /* Start a scheduled AOF rewrite if this was requested by the user while
     * a BGSAVE was in progress. */
A
antirez 已提交
852
    if (server.rdb_child_pid == -1 && server.aof_child_pid == -1 &&
853
        server.aof_rewrite_scheduled)
854 855 856 857
    {
        rewriteAppendOnlyFileBackground();
    }

A
antirez 已提交
858
    /* Check if a background saving or AOF rewrite in progress terminated. */
A
antirez 已提交
859
    if (server.rdb_child_pid != -1 || server.aof_child_pid != -1) {
860 861 862 863
        int statloc;
        pid_t pid;

        if ((pid = wait3(&statloc,WNOHANG,NULL)) != 0) {
864 865 866 867 868
            int exitcode = WEXITSTATUS(statloc);
            int bysignal = 0;
            
            if (WIFSIGNALED(statloc)) bysignal = WTERMSIG(statloc);

A
antirez 已提交
869
            if (pid == server.rdb_child_pid) {
870
                backgroundSaveDoneHandler(exitcode,bysignal);
871
            } else {
872
                backgroundRewriteDoneHandler(exitcode,bysignal);
873 874 875
            }
            updateDictResizePolicy();
        }
A
antirez 已提交
876
    } else {
877 878
        /* If there is not a background saving/rewrite in progress check if
         * we have to save/rewrite now */
879 880 881 882
         for (j = 0; j < server.saveparamslen; j++) {
            struct saveparam *sp = server.saveparams+j;

            if (server.dirty >= sp->changes &&
883
                server.unixtime-server.lastsave > sp->seconds) {
884 885
                redisLog(REDIS_NOTICE,"%d changes in %d seconds. Saving...",
                    sp->changes, sp->seconds);
A
antirez 已提交
886
                rdbSaveBackground(server.rdb_filename);
887 888 889
                break;
            }
         }
890 891

         /* Trigger an AOF rewrite if needed */
A
antirez 已提交
892
         if (server.rdb_child_pid == -1 &&
A
antirez 已提交
893
             server.aof_child_pid == -1 &&
894 895
             server.aof_rewrite_perc &&
             server.aof_current_size > server.aof_rewrite_min_size)
896
         {
897 898 899 900
            long long base = server.aof_rewrite_base_size ?
                            server.aof_rewrite_base_size : 1;
            long long growth = (server.aof_current_size*100/base) - 100;
            if (growth >= server.aof_rewrite_perc) {
901
                redisLog(REDIS_NOTICE,"Starting automatic rewriting of AOF on %lld%% growth",growth);
902 903 904
                rewriteAppendOnlyFileBackground();
            }
         }
905 906
    }

907 908 909 910 911

    /* If we postponed an AOF buffer flush, let's try to do it every time the
     * cron function is called. */
    if (server.aof_flush_postponed_start) flushAppendOnlyFile(0);

912 913 914 915
    /* Expire a few keys per cycle, only if this is a master.
     * On slaves we wait for DEL operations synthesized by the master
     * in order to guarantee a strict consistency. */
    if (server.masterhost == NULL) activeExpireCycle();
916

917 918 919
    /* Close clients that need to be closed asynchronous */
    freeClientsInAsyncFreeQueue();

920 921
    /* Replication cron function -- used to reconnect to master and
     * to detect transfer failures. */
922
    if (!(loops % 10)) replicationCron();
923

A
antirez 已提交
924 925 926
    /* Run other sub-systems specific cron jobs */
    if (server.cluster_enabled && !(loops % 10)) clusterCron();

927
    server.cronloops++;
928 929 930 931 932 933 934 935
    return 100;
}

/* This function gets called every time Redis is entering the
 * main loop of the event driven library, that is, before to sleep
 * for ready file descriptors. */
void beforeSleep(struct aeEventLoop *eventLoop) {
    REDIS_NOTUSED(eventLoop);
936 937
    listNode *ln;
    redisClient *c;
938

939 940 941 942 943 944
    /* Try to process pending commands for clients that were just unblocked. */
    while (listLength(server.unblocked_clients)) {
        ln = listFirst(server.unblocked_clients);
        redisAssert(ln != NULL);
        c = ln->value;
        listDelNode(server.unblocked_clients,ln);
945
        c->flags &= ~REDIS_UNBLOCKED;
946 947

        /* Process remaining data in the input buffer. */
948 949
        if (c->querybuf && sdslen(c->querybuf) > 0) {
            server.current_client = c;
950
            processInputBuffer(c);
951 952
            server.current_client = NULL;
        }
953 954
    }

955
    /* Write the AOF buffer on disk */
956
    flushAppendOnlyFile(0);
957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985
}

/* =========================== Server initialization ======================== */

void createSharedObjects(void) {
    int j;

    shared.crlf = createObject(REDIS_STRING,sdsnew("\r\n"));
    shared.ok = createObject(REDIS_STRING,sdsnew("+OK\r\n"));
    shared.err = createObject(REDIS_STRING,sdsnew("-ERR\r\n"));
    shared.emptybulk = createObject(REDIS_STRING,sdsnew("$0\r\n\r\n"));
    shared.czero = createObject(REDIS_STRING,sdsnew(":0\r\n"));
    shared.cone = createObject(REDIS_STRING,sdsnew(":1\r\n"));
    shared.cnegone = createObject(REDIS_STRING,sdsnew(":-1\r\n"));
    shared.nullbulk = createObject(REDIS_STRING,sdsnew("$-1\r\n"));
    shared.nullmultibulk = createObject(REDIS_STRING,sdsnew("*-1\r\n"));
    shared.emptymultibulk = createObject(REDIS_STRING,sdsnew("*0\r\n"));
    shared.pong = createObject(REDIS_STRING,sdsnew("+PONG\r\n"));
    shared.queued = createObject(REDIS_STRING,sdsnew("+QUEUED\r\n"));
    shared.wrongtypeerr = createObject(REDIS_STRING,sdsnew(
        "-ERR Operation against a key holding the wrong kind of value\r\n"));
    shared.nokeyerr = createObject(REDIS_STRING,sdsnew(
        "-ERR no such key\r\n"));
    shared.syntaxerr = createObject(REDIS_STRING,sdsnew(
        "-ERR syntax error\r\n"));
    shared.sameobjecterr = createObject(REDIS_STRING,sdsnew(
        "-ERR source and destination objects are the same\r\n"));
    shared.outofrangeerr = createObject(REDIS_STRING,sdsnew(
        "-ERR index out of range\r\n"));
A
antirez 已提交
986 987
    shared.noscripterr = createObject(REDIS_STRING,sdsnew(
        "-NOSCRIPT No matching script. Please use EVAL.\r\n"));
988 989
    shared.loadingerr = createObject(REDIS_STRING,sdsnew(
        "-LOADING Redis is loading the dataset in memory\r\n"));
990
    shared.slowscripterr = createObject(REDIS_STRING,sdsnew(
991
        "-BUSY Redis is busy running a script. You can only call SCRIPT KILL or SHUTDOWN NOSAVE.\r\n"));
992
    shared.bgsaveerr = createObject(REDIS_STRING,sdsnew(
993 994 995
        "-MISCONF Redis is configured to save RDB snapshots, but is currently not able to persist on disk. Commands that may modify the data set are disabled. Please check Redis logs for details about the error.\r\n"));
    shared.roslaveerr = createObject(REDIS_STRING,sdsnew(
        "-READONLY You can't write against a read only slave.\r\n"));
996 997
    shared.oomerr = createObject(REDIS_STRING,sdsnew(
        "-OOM command not allowed when used memory > 'maxmemory'.\r\n"));
998 999 1000
    shared.space = createObject(REDIS_STRING,sdsnew(" "));
    shared.colon = createObject(REDIS_STRING,sdsnew(":"));
    shared.plus = createObject(REDIS_STRING,sdsnew("+"));
1001 1002 1003 1004 1005

    for (j = 0; j < REDIS_SHARED_SELECT_CMDS; j++) {
        shared.select[j] = createObject(REDIS_STRING,
            sdscatprintf(sdsempty(),"select %d\r\n", j));
    }
1006 1007 1008 1009 1010 1011
    shared.messagebulk = createStringObject("$7\r\nmessage\r\n",13);
    shared.pmessagebulk = createStringObject("$8\r\npmessage\r\n",14);
    shared.subscribebulk = createStringObject("$9\r\nsubscribe\r\n",15);
    shared.unsubscribebulk = createStringObject("$11\r\nunsubscribe\r\n",18);
    shared.psubscribebulk = createStringObject("$10\r\npsubscribe\r\n",17);
    shared.punsubscribebulk = createStringObject("$12\r\npunsubscribe\r\n",19);
1012
    shared.del = createStringObject("DEL",3);
1013 1014
    shared.rpop = createStringObject("RPOP",4);
    shared.lpop = createStringObject("LPOP",4);
1015 1016 1017 1018
    for (j = 0; j < REDIS_SHARED_INTEGERS; j++) {
        shared.integers[j] = createObject(REDIS_STRING,(void*)(long)j);
        shared.integers[j]->encoding = REDIS_ENCODING_INT;
    }
1019 1020 1021 1022 1023 1024
    for (j = 0; j < REDIS_SHARED_BULKHDR_LEN; j++) {
        shared.mbulkhdr[j] = createObject(REDIS_STRING,
            sdscatprintf(sdsempty(),"*%d\r\n",j));
        shared.bulkhdr[j] = createObject(REDIS_STRING,
            sdscatprintf(sdsempty(),"$%d\r\n",j));
    }
1025 1026 1027
}

void initServerConfig() {
A
antirez 已提交
1028 1029
    getRandomHexChars(server.runid,REDIS_RUN_ID_SIZE);
    server.runid[REDIS_RUN_ID_SIZE] = '\0';
1030
    server.arch_bits = (sizeof(long) == 8) ? 64 : 32;
1031
    server.port = REDIS_SERVERPORT;
1032
    server.bindaddr = NULL;
1033
    server.unixsocket = NULL;
1034
    server.unixsocketperm = 0;
1035 1036 1037
    server.ipfd = -1;
    server.sofd = -1;
    server.dbnum = REDIS_DEFAULT_DBNUM;
A
antirez 已提交
1038
    server.verbosity = REDIS_NOTICE;
1039
    server.maxidletime = REDIS_MAXIDLETIME;
1040
    server.client_max_querybuf_len = REDIS_MAX_QUERYBUF_LEN;
1041
    server.saveparams = NULL;
1042
    server.loading = 0;
1043
    server.logfile = NULL; /* NULL = log on standard output */
J
Jonah H. Harris 已提交
1044 1045 1046
    server.syslog_enabled = 0;
    server.syslog_ident = zstrdup("redis");
    server.syslog_facility = LOG_LOCAL0;
1047
    server.daemonize = 0;
1048
    server.aof_state = REDIS_AOF_OFF;
1049 1050 1051 1052 1053 1054
    server.aof_fsync = AOF_FSYNC_EVERYSEC;
    server.aof_no_fsync_on_rewrite = 0;
    server.aof_rewrite_perc = REDIS_AOF_REWRITE_PERC;
    server.aof_rewrite_min_size = REDIS_AOF_REWRITE_MIN_SIZE;
    server.aof_rewrite_base_size = 0;
    server.aof_rewrite_scheduled = 0;
A
antirez 已提交
1055
    server.aof_last_fsync = time(NULL);
1056
    server.aof_delayed_fsync = 0;
A
antirez 已提交
1057 1058
    server.aof_fd = -1;
    server.aof_selected_db = -1; /* Make sure the first time will not match */
1059
    server.aof_flush_postponed_start = 0;
1060
    server.pidfile = zstrdup("/var/run/redis.pid");
A
antirez 已提交
1061
    server.rdb_filename = zstrdup("dump.rdb");
1062
    server.aof_filename = zstrdup("appendonly.aof");
1063
    server.requirepass = NULL;
A
antirez 已提交
1064
    server.rdb_compression = 1;
1065
    server.rdb_checksum = 1;
1066
    server.activerehashing = 1;
1067
    server.maxclients = REDIS_MAX_CLIENTS;
1068
    server.bpop_blocked_clients = 0;
1069
    server.maxmemory = 0;
1070 1071
    server.maxmemory_policy = REDIS_MAXMEMORY_VOLATILE_LRU;
    server.maxmemory_samples = 3;
1072 1073
    server.hash_max_ziplist_entries = REDIS_HASH_MAX_ZIPLIST_ENTRIES;
    server.hash_max_ziplist_value = REDIS_HASH_MAX_ZIPLIST_VALUE;
1074 1075
    server.list_max_ziplist_entries = REDIS_LIST_MAX_ZIPLIST_ENTRIES;
    server.list_max_ziplist_value = REDIS_LIST_MAX_ZIPLIST_VALUE;
1076
    server.set_max_intset_entries = REDIS_SET_MAX_INTSET_ENTRIES;
1077 1078
    server.zset_max_ziplist_entries = REDIS_ZSET_MAX_ZIPLIST_ENTRIES;
    server.zset_max_ziplist_value = REDIS_ZSET_MAX_ZIPLIST_VALUE;
1079
    server.shutdown_asap = 0;
A
1fe4cd5  
antirez 已提交
1080 1081
    server.repl_ping_slave_period = REDIS_REPL_PING_SLAVE_PERIOD;
    server.repl_timeout = REDIS_REPL_TIMEOUT;
A
antirez 已提交
1082
    server.cluster_enabled = 0;
1083
    server.cluster.configfile = zstrdup("nodes.conf");
1084
    server.lua_caller = NULL;
A
antirez 已提交
1085
    server.lua_time_limit = REDIS_LUA_TIME_LIMIT;
1086
    server.lua_client = NULL;
1087
    server.lua_timedout = 0;
1088

A
antirez 已提交
1089
    updateLRUClock();
1090 1091 1092 1093 1094 1095 1096 1097 1098 1099
    resetServerSaveParams();

    appendServerSaveParams(60*60,1);  /* save after 1 hour and 1 change */
    appendServerSaveParams(300,100);  /* save after 5 minutes and 100 changes */
    appendServerSaveParams(60,10000); /* save after 1 minute and 10000 changes */
    /* Replication related */
    server.masterauth = NULL;
    server.masterhost = NULL;
    server.masterport = 6379;
    server.master = NULL;
A
antirez 已提交
1100
    server.repl_state = REDIS_REPL_NONE;
1101
    server.repl_syncio_timeout = REDIS_REPL_SYNCIO_TIMEOUT;
1102
    server.repl_serve_stale_data = 1;
1103
    server.repl_slave_ro = 1;
1104
    server.repl_down_since = time(NULL);
1105

1106 1107 1108 1109
    /* Client output buffer limits */
    server.client_obuf_limits[REDIS_CLIENT_LIMIT_CLASS_NORMAL].hard_limit_bytes = 0;
    server.client_obuf_limits[REDIS_CLIENT_LIMIT_CLASS_NORMAL].soft_limit_bytes = 0;
    server.client_obuf_limits[REDIS_CLIENT_LIMIT_CLASS_NORMAL].soft_limit_seconds = 0;
1110 1111
    server.client_obuf_limits[REDIS_CLIENT_LIMIT_CLASS_SLAVE].hard_limit_bytes = 1024*1024*256;
    server.client_obuf_limits[REDIS_CLIENT_LIMIT_CLASS_SLAVE].soft_limit_bytes = 1024*1024*64;
1112
    server.client_obuf_limits[REDIS_CLIENT_LIMIT_CLASS_SLAVE].soft_limit_seconds = 60;
1113 1114
    server.client_obuf_limits[REDIS_CLIENT_LIMIT_CLASS_PUBSUB].hard_limit_bytes = 1024*1024*32;
    server.client_obuf_limits[REDIS_CLIENT_LIMIT_CLASS_PUBSUB].soft_limit_bytes = 1024*1024*8;
1115 1116
    server.client_obuf_limits[REDIS_CLIENT_LIMIT_CLASS_PUBSUB].soft_limit_seconds = 60;

1117 1118 1119 1120 1121
    /* Double constants initialization */
    R_Zero = 0.0;
    R_PosInf = 1.0/R_Zero;
    R_NegInf = -1.0/R_Zero;
    R_Nan = R_Zero/R_Zero;
1122 1123 1124 1125 1126 1127 1128 1129

    /* Command table -- we intiialize it here as it is part of the
     * initial configuration, since command names may be changed via
     * redis.conf using the rename-command directive. */
    server.commands = dictCreate(&commandTableDictType,NULL);
    populateCommandTable();
    server.delCommand = lookupCommandByCString("del");
    server.multiCommand = lookupCommandByCString("multi");
1130
    server.lpushCommand = lookupCommandByCString("lpush");
1131 1132 1133 1134
    
    /* Slow log */
    server.slowlog_log_slower_than = REDIS_SLOWLOG_LOG_SLOWER_THAN;
    server.slowlog_max_len = REDIS_SLOWLOG_MAX_LEN;
A
antirez 已提交
1135

A
antirez 已提交
1136
    /* Debugging */
A
antirez 已提交
1137 1138 1139 1140
    server.assert_failed = "<no assertion failed>";
    server.assert_file = "<no file>";
    server.assert_line = 0;
    server.bug_report_start = 0;
A
antirez 已提交
1141
    server.watchdog_period = 0;
1142 1143
}

1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165
/* This function will try to raise the max number of open files accordingly to
 * the configured max number of clients. It will also account for 32 additional
 * file descriptors as we need a few more for persistence, listening
 * sockets, log files and so forth.
 *
 * If it will not be possible to set the limit accordingly to the configured
 * max number of clients, the function will do the reverse setting
 * server.maxclients to the value that we can actually handle. */
void adjustOpenFilesLimit(void) {
    rlim_t maxfiles = server.maxclients+32;
    struct rlimit limit;

    if (getrlimit(RLIMIT_NOFILE,&limit) == -1) {
        redisLog(REDIS_WARNING,"Unable to obtain the current NOFILE limit (%s), assuming 1024 and setting the max clients configuration accordingly.",
            strerror(errno));
        server.maxclients = 1024-32;
    } else {
        rlim_t oldlimit = limit.rlim_cur;

        /* Set the max number of files if the current limit is not enough
         * for our needs. */
        if (oldlimit < maxfiles) {
1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177
            rlim_t f;
            
            f = maxfiles;
            while(f > oldlimit) {
                limit.rlim_cur = f;
                limit.rlim_max = f;
                if (setrlimit(RLIMIT_NOFILE,&limit) != -1) break;
                f -= 128;
            }
            if (f < oldlimit) f = oldlimit;
            if (f != maxfiles) {
                server.maxclients = f-32;
1178 1179 1180 1181 1182 1183 1184 1185 1186 1187
                redisLog(REDIS_WARNING,"Unable to set the max number of files limit to %d (%s), setting the max clients configuration to %d.",
                    (int) maxfiles, strerror(errno), (int) server.maxclients);
            } else {
                redisLog(REDIS_NOTICE,"Max number of open files set to %d",
                    (int) maxfiles);
            }
        }
    }
}

1188 1189 1190 1191 1192
void initServer() {
    int j;

    signal(SIGHUP, SIG_IGN);
    signal(SIGPIPE, SIG_IGN);
1193
    setupSignalHandlers();
1194

J
Jonah H. Harris 已提交
1195 1196 1197 1198 1199
    if (server.syslog_enabled) {
        openlog(server.syslog_ident, LOG_PID | LOG_NDELAY | LOG_NOWAIT,
            server.syslog_facility);
    }

1200
    server.current_client = NULL;
1201
    server.clients = listCreate();
1202
    server.clients_to_close = listCreate();
1203 1204
    server.slaves = listCreate();
    server.monitors = listCreate();
1205
    server.unblocked_clients = listCreate();
1206

1207
    createSharedObjects();
1208 1209
    adjustOpenFilesLimit();
    server.el = aeCreateEventLoop(server.maxclients+1024);
1210
    server.db = zmalloc(sizeof(redisDb)*server.dbnum);
1211

1212
    if (server.port != 0) {
1213
        server.ipfd = anetTcpServer(server.neterr,server.port,server.bindaddr);
1214
        if (server.ipfd == ANET_ERR) {
1215 1216
            redisLog(REDIS_WARNING, "Opening port %d: %s",
                server.port, server.neterr);
1217 1218
            exit(1);
        }
1219
    }
1220 1221
    if (server.unixsocket != NULL) {
        unlink(server.unixsocket); /* don't care if this fails */
1222
        server.sofd = anetUnixServer(server.neterr,server.unixsocket,server.unixsocketperm);
1223 1224 1225 1226
        if (server.sofd == ANET_ERR) {
            redisLog(REDIS_WARNING, "Opening socket: %s", server.neterr);
            exit(1);
        }
1227
    }
1228 1229
    if (server.ipfd < 0 && server.sofd < 0) {
        redisLog(REDIS_WARNING, "Configured to not listen anywhere, exiting.");
1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243
        exit(1);
    }
    for (j = 0; j < server.dbnum; j++) {
        server.db[j].dict = dictCreate(&dbDictType,NULL);
        server.db[j].expires = dictCreate(&keyptrDictType,NULL);
        server.db[j].blocking_keys = dictCreate(&keylistDictType,NULL);
        server.db[j].watched_keys = dictCreate(&keylistDictType,NULL);
        server.db[j].id = j;
    }
    server.pubsub_channels = dictCreate(&keylistDictType,NULL);
    server.pubsub_patterns = listCreate();
    listSetFreeMethod(server.pubsub_patterns,freePubsubPattern);
    listSetMatchMethod(server.pubsub_patterns,listMatchPubsubPattern);
    server.cronloops = 0;
A
antirez 已提交
1244
    server.rdb_child_pid = -1;
A
antirez 已提交
1245 1246 1247
    server.aof_child_pid = -1;
    server.aof_rewrite_buf = sdsempty();
    server.aof_buf = sdsempty();
1248 1249 1250 1251 1252
    server.lastsave = time(NULL);
    server.dirty = 0;
    server.stat_numcommands = 0;
    server.stat_numconnections = 0;
    server.stat_expiredkeys = 0;
1253
    server.stat_evictedkeys = 0;
1254
    server.stat_starttime = time(NULL);
1255 1256
    server.stat_keyspace_misses = 0;
    server.stat_keyspace_hits = 0;
1257
    server.stat_peak_memory = 0;
1258
    server.stat_fork_time = 0;
1259
    server.stat_rejected_conn = 0;
1260 1261 1262 1263
    memset(server.ops_sec_samples,0,sizeof(server.ops_sec_samples));
    server.ops_sec_idx = 0;
    server.ops_sec_last_sample_time = mstime();
    server.ops_sec_last_sample_ops = 0;
1264
    server.unixtime = time(NULL);
1265
    server.lastbgsave_status = REDIS_OK;
1266
    server.stop_writes_on_bgsave_err = 1;
1267
    aeCreateTimeEvent(server.el, 1, serverCron, NULL, NULL);
1268
    if (server.ipfd > 0 && aeCreateFileEvent(server.el,server.ipfd,AE_READABLE,
1269
        acceptTcpHandler,NULL) == AE_ERR) oom("creating file event");
1270
    if (server.sofd > 0 && aeCreateFileEvent(server.el,server.sofd,AE_READABLE,
1271
        acceptUnixHandler,NULL) == AE_ERR) oom("creating file event");
1272

1273
    if (server.aof_state == REDIS_AOF_ON) {
A
antirez 已提交
1274
        server.aof_fd = open(server.aof_filename,
1275
                               O_WRONLY|O_APPEND|O_CREAT,0644);
A
antirez 已提交
1276
        if (server.aof_fd == -1) {
1277 1278 1279 1280 1281 1282
            redisLog(REDIS_WARNING, "Can't open the append-only file: %s",
                strerror(errno));
            exit(1);
        }
    }

1283 1284 1285 1286 1287 1288 1289 1290 1291 1292
    /* 32 bit instances are limited to 4GB of address space, so if there is
     * no explicit limit in the user provided configuration we set a limit
     * at 3.5GB using maxmemory with 'noeviction' policy'. This saves
     * useless crashes of the Redis instance. */
    if (server.arch_bits == 32 && server.maxmemory == 0) {
        redisLog(REDIS_WARNING,"Warning: 32 bit instance detected but no memory limit set. Setting 3.5 GB maxmemory limit with 'noeviction' policy now.");
        server.maxmemory = 3584LL*(1024*1024); /* 3584 MB = 3.5 GB */
        server.maxmemory_policy = REDIS_MAXMEMORY_NO_EVICTION;
    }

A
antirez 已提交
1293
    if (server.cluster_enabled) clusterInit();
1294
    scriptingInit();
1295
    slowlogInit();
1296
    bioInit();
1297 1298
}

1299 1300 1301 1302
/* Populates the Redis Command Table starting from the hard coded list
 * we have on top of redis.c file. */
void populateCommandTable(void) {
    int j;
1303
    int numcommands = sizeof(redisCommandTable)/sizeof(struct redisCommand);
1304 1305

    for (j = 0; j < numcommands; j++) {
1306
        struct redisCommand *c = redisCommandTable+j;
1307
        char *f = c->sflags;
1308
        int retval;
1309

1310 1311 1312 1313 1314 1315 1316 1317
        while(*f != '\0') {
            switch(*f) {
            case 'w': c->flags |= REDIS_CMD_WRITE; break;
            case 'r': c->flags |= REDIS_CMD_READONLY; break;
            case 'm': c->flags |= REDIS_CMD_DENYOOM; break;
            case 'a': c->flags |= REDIS_CMD_ADMIN; break;
            case 'p': c->flags |= REDIS_CMD_PUBSUB; break;
            case 'f': c->flags |= REDIS_CMD_FORCE_REPLICATION; break;
1318 1319
            case 's': c->flags |= REDIS_CMD_NOSCRIPT; break;
            case 'R': c->flags |= REDIS_CMD_RANDOM; break;
1320
            case 'S': c->flags |= REDIS_CMD_SORT_FOR_SCRIPT; break;
1321 1322 1323 1324 1325
            default: redisPanic("Unsupported command flag"); break;
            }
            f++;
        }

1326 1327 1328
        retval = dictAdd(server.commands, sdsnew(c->name), c);
        assert(retval == DICT_OK);
    }
1329 1330
}

1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342
void resetCommandTableStats(void) {
    int numcommands = sizeof(redisCommandTable)/sizeof(struct redisCommand);
    int j;

    for (j = 0; j < numcommands; j++) {
        struct redisCommand *c = redisCommandTable+j;

        c->microseconds = 0;
        c->calls = 0;
    }
}

1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379
/* ========================== Redis OP Array API ============================ */

void redisOpArrayInit(redisOpArray *oa) {
    oa->ops = NULL;
    oa->numops = 0;
}

int redisOpArrayAppend(redisOpArray *oa, struct redisCommand *cmd, int dbid,
                       robj **argv, int argc, int target)
{
    redisOp *op;

    oa->ops = zrealloc(oa->ops,sizeof(redisOp)*(oa->numops+1));
    op = oa->ops+oa->numops;
    op->cmd = cmd;
    op->dbid = dbid;
    op->argv = argv;
    op->argc = argc;
    op->target = target;
    oa->numops++;
    return oa->numops;
}

void redisOpArrayFree(redisOpArray *oa) {
    while(oa->numops) {
        int j;
        redisOp *op;

        oa->numops--;
        op = oa->ops+oa->numops;
        for (j = 0; j < op->argc; j++)
            decrRefCount(op->argv[j]);
        zfree(op->argv);
    }
    zfree(oa->ops);
}

1380 1381
/* ====================== Commands lookup and execution ===================== */

1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392
struct redisCommand *lookupCommand(sds name) {
    return dictFetchValue(server.commands, name);
}

struct redisCommand *lookupCommandByCString(char *s) {
    struct redisCommand *cmd;
    sds name = sdsnew(s);

    cmd = dictFetchValue(server.commands, name);
    sdsfree(name);
    return cmd;
1393 1394
}

A
antirez 已提交
1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411
/* Propagate the specified command (in the context of the specified database id)
 * to AOF, Slaves and Monitors.
 *
 * flags are an xor between:
 * + REDIS_PROPAGATE_NONE (no propagation of command at all)
 * + REDIS_PROPAGATE_AOF (propagate into the AOF file if is enabled)
 * + REDIS_PROPAGATE_REPL (propagate into the replication link)
 */
void propagate(struct redisCommand *cmd, int dbid, robj **argv, int argc,
               int flags)
{
    if (server.aof_state != REDIS_AOF_OFF && flags & REDIS_PROPAGATE_AOF)
        feedAppendOnlyFile(cmd,dbid,argv,argc);
    if (flags & REDIS_PROPAGATE_REPL && listLength(server.slaves))
        replicationFeedSlaves(server.slaves,dbid,argv,argc);
}

1412 1413
/* Used inside commands to schedule the propagation of additional commands
 * after the current command is propagated to AOF / Replication. */
1414 1415 1416
void alsoPropagate(struct redisCommand *cmd, int dbid, robj **argv, int argc,
                   int target)
{
1417
    redisOpArrayAppend(&server.also_propagate,cmd,dbid,argv,argc,target);
1418 1419
}

1420
/* Call() is the core of Redis execution of a command */
1421
void call(redisClient *c, int flags) {
1422
    long long dirty, start = ustime(), duration;
1423

A
antirez 已提交
1424 1425 1426
    /* Sent the command to clients in MONITOR mode, only if the commands are
     * not geneated from reading an AOF. */
    if (listLength(server.monitors) && !server.loading)
1427
        replicationFeedMonitors(c,server.monitors,c->db->id,c->argv,c->argc);
A
antirez 已提交
1428 1429

    /* Call the command. */
1430
    redisOpArrayInit(&server.also_propagate);
1431
    dirty = server.dirty;
1432
    c->cmd->proc(c);
1433
    dirty = server.dirty-dirty;
1434
    duration = ustime()-start;
1435 1436 1437 1438 1439 1440

    /* When EVAL is called loading the AOF we don't want commands called
     * from Lua to go into the slowlog or to populate statistics. */
    if (server.loading && c->flags & REDIS_LUA_CLIENT)
        flags &= ~(REDIS_CALL_SLOWLOG | REDIS_CALL_STATS);

A
antirez 已提交
1441 1442
    /* Log the command into the Slow log if needed, and populate the
     * per-command statistics that we show in INFO commandstats. */
1443 1444 1445 1446 1447 1448
    if (flags & REDIS_CALL_SLOWLOG)
        slowlogPushEntryIfNeeded(c->argv,c->argc,duration);
    if (flags & REDIS_CALL_STATS) {
        c->cmd->microseconds += duration;
        c->cmd->calls++;
    }
A
antirez 已提交
1449 1450

    /* Propagate the command into the AOF and replication link */
1451
    if (flags & REDIS_CALL_PROPAGATE) {
A
antirez 已提交
1452 1453 1454 1455 1456 1457 1458 1459
        int flags = REDIS_PROPAGATE_NONE;

        if (c->cmd->flags & REDIS_CMD_FORCE_REPLICATION)
            flags |= REDIS_PROPAGATE_REPL;
        if (dirty)
            flags |= (REDIS_PROPAGATE_REPL | REDIS_PROPAGATE_AOF);
        if (flags != REDIS_PROPAGATE_NONE)
            propagate(c->cmd,c->db->id,c->argv,c->argc,flags);
1460
    }
1461 1462
    /* Commands such as LPUSH or BRPOPLPUSH may propagate an additional
     * PUSH command. */
1463
    if (server.also_propagate.numops) {
1464
        int j;
1465
        redisOp *rop;
1466

1467 1468 1469 1470 1471
        for (j = 0; j < server.also_propagate.numops; j++) {
            rop = &server.also_propagate.ops[j];
            propagate(rop->cmd, rop->dbid, rop->argv, rop->argc, rop->target);
        }
        redisOpArrayFree(&server.also_propagate);
1472
    }
1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484
    server.stat_numcommands++;
}

/* If this function gets called we already read a whole
 * command, argments are in the client argv/argc fields.
 * processCommand() execute the command or prepare the
 * server for a bulk read from the client.
 *
 * If 1 is returned the client is still alive and valid and
 * and other operations can be performed by the caller. Otherwise
 * if 0 is returned the client was destroied (i.e. after QUIT). */
int processCommand(redisClient *c) {
P
Pieter Noordhuis 已提交
1485 1486 1487 1488
    /* The QUIT command is handled separately. Normal command procs will
     * go through checking for replication and QUIT will cause trouble
     * when FORCE_REPLICATION is enabled and would be implemented in
     * a regular command proc. */
1489
    if (!strcasecmp(c->argv[0]->ptr,"quit")) {
P
Pieter Noordhuis 已提交
1490
        addReply(c,shared.ok);
1491
        c->flags |= REDIS_CLOSE_AFTER_REPLY;
1492
        return REDIS_ERR;
1493 1494 1495
    }

    /* Now lookup the command and check ASAP about trivial error conditions
1496
     * such as wrong arity, bad command name and so forth. */
1497
    c->cmd = c->lastcmd = lookupCommand(c->argv[0]->ptr);
1498
    if (!c->cmd) {
1499 1500
        addReplyErrorFormat(c,"unknown command '%s'",
            (char*)c->argv[0]->ptr);
1501
        return REDIS_OK;
1502 1503
    } else if ((c->cmd->arity > 0 && c->cmd->arity != c->argc) ||
               (c->argc < -c->cmd->arity)) {
1504
        addReplyErrorFormat(c,"wrong number of arguments for '%s' command",
1505
            c->cmd->name);
1506
        return REDIS_OK;
1507 1508 1509
    }

    /* Check if the user is authenticated */
1510 1511
    if (server.requirepass && !c->authenticated && c->cmd->proc != authCommand)
    {
1512
        addReplyError(c,"operation not permitted");
1513
        return REDIS_OK;
1514 1515
    }

A
antirez 已提交
1516 1517
    /* If cluster is enabled, redirect here */
    if (server.cluster_enabled &&
1518
                !(c->cmd->getkeys_proc == NULL && c->cmd->firstkey == 0)) {
A
antirez 已提交
1519 1520 1521 1522 1523 1524
        int hashslot;

        if (server.cluster.state != REDIS_CLUSTER_OK) {
            addReplyError(c,"The cluster is down. Check with CLUSTER INFO for more information");
            return REDIS_OK;
        } else {
1525
            int ask;
1526
            clusterNode *n = getNodeByQuery(c,c->cmd,c->argv,c->argc,&hashslot,&ask);
A
antirez 已提交
1527
            if (n == NULL) {
1528
                addReplyError(c,"Multi keys request invalid in cluster");
A
antirez 已提交
1529 1530 1531
                return REDIS_OK;
            } else if (n != server.cluster.myself) {
                addReplySds(c,sdscatprintf(sdsempty(),
1532 1533
                    "-%s %d %s:%d\r\n", ask ? "ASK" : "MOVED",
                    hashslot,n->ip,n->port));
A
antirez 已提交
1534 1535 1536 1537 1538
                return REDIS_OK;
            }
        }
    }

1539 1540 1541 1542 1543
    /* Handle the maxmemory directive.
     *
     * First we try to free some memory if possible (if there are volatile
     * keys in the dataset). If there are not the only thing we can do
     * is returning an error. */
1544 1545 1546
    if (server.maxmemory) {
        int retval = freeMemoryIfNeeded();
        if ((c->cmd->flags & REDIS_CMD_DENYOOM) && retval == REDIS_ERR) {
1547
            addReply(c, shared.oomerr);
1548 1549
            return REDIS_OK;
        }
1550 1551
    }

1552
    /* Don't accept write commands if there are problems persisting on disk. */
1553 1554 1555
    if (server.stop_writes_on_bgsave_err &&
        server.saveparamslen > 0
        && server.lastbgsave_status == REDIS_ERR &&
1556 1557 1558 1559 1560 1561
        c->cmd->flags & REDIS_CMD_WRITE)
    {
        addReply(c, shared.bgsaveerr);
        return REDIS_OK;
    }

1562 1563 1564 1565 1566 1567 1568 1569 1570 1571
    /* Don't accept wirte commands if this is a read only slave. But
     * accept write commands if this is our master. */
    if (server.masterhost && server.repl_slave_ro &&
        !(c->flags & REDIS_MASTER) &&
        c->cmd->flags & REDIS_CMD_WRITE)
    {
        addReply(c, shared.roslaveerr);
        return REDIS_OK;
    }

1572 1573 1574
    /* Only allow SUBSCRIBE and UNSUBSCRIBE in the context of Pub/Sub */
    if ((dictSize(c->pubsub_channels) > 0 || listLength(c->pubsub_patterns) > 0)
        &&
1575 1576 1577 1578
        c->cmd->proc != subscribeCommand &&
        c->cmd->proc != unsubscribeCommand &&
        c->cmd->proc != psubscribeCommand &&
        c->cmd->proc != punsubscribeCommand) {
1579
        addReplyError(c,"only (P)SUBSCRIBE / (P)UNSUBSCRIBE / QUIT allowed in this context");
1580
        return REDIS_OK;
1581 1582
    }

1583 1584
    /* Only allow INFO and SLAVEOF when slave-serve-stale-data is no and
     * we are a slave with a broken link with master. */
A
antirez 已提交
1585
    if (server.masterhost && server.repl_state != REDIS_REPL_CONNECTED &&
1586
        server.repl_serve_stale_data == 0 &&
1587
        c->cmd->proc != infoCommand && c->cmd->proc != slaveofCommand)
1588 1589 1590 1591 1592 1593
    {
        addReplyError(c,
            "link with MASTER is down and slave-serve-stale-data is set to no");
        return REDIS_OK;
    }

1594
    /* Loading DB? Return an error if the command is not INFO */
1595
    if (server.loading && c->cmd->proc != infoCommand) {
1596 1597 1598 1599
        addReply(c, shared.loadingerr);
        return REDIS_OK;
    }

1600 1601
    /* Lua script too slow? Only allow SHUTDOWN NOSAVE and SCRIPT KILL. */
    if (server.lua_timedout &&
1602
        !(c->cmd->proc == shutdownCommand &&
1603 1604 1605 1606 1607 1608
          c->argc == 2 &&
          tolower(((char*)c->argv[1]->ptr)[0]) == 'n') &&
        !(c->cmd->proc == scriptCommand &&
          c->argc == 2 &&
          tolower(((char*)c->argv[1]->ptr)[0]) == 'k'))
    {
1609 1610 1611 1612
        addReply(c, shared.slowscripterr);
        return REDIS_OK;
    }

1613 1614
    /* Exec the command */
    if (c->flags & REDIS_MULTI &&
1615 1616
        c->cmd->proc != execCommand && c->cmd->proc != discardCommand &&
        c->cmd->proc != multiCommand && c->cmd->proc != watchCommand)
1617
    {
1618
        queueMultiCommand(c);
1619 1620
        addReply(c,shared.queued);
    } else {
1621
        call(c,REDIS_CALL_FULL);
1622
    }
1623
    return REDIS_OK;
1624 1625 1626 1627
}

/*================================== Shutdown =============================== */

1628 1629 1630 1631
int prepareForShutdown(int flags) {
    int save = flags & REDIS_SHUTDOWN_SAVE;
    int nosave = flags & REDIS_SHUTDOWN_NOSAVE;

1632
    redisLog(REDIS_WARNING,"User requested shutdown...");
1633 1634 1635
    /* Kill the saving child if there is a background saving in progress.
       We want to avoid race conditions, for instance our saving child may
       overwrite the synchronous saving did by SHUTDOWN. */
A
antirez 已提交
1636
    if (server.rdb_child_pid != -1) {
1637
        redisLog(REDIS_WARNING,"There is a child saving an .rdb. Killing it!");
A
antirez 已提交
1638 1639
        kill(server.rdb_child_pid,SIGKILL);
        rdbRemoveTempFile(server.rdb_child_pid);
1640
    }
1641
    if (server.aof_state != REDIS_AOF_OFF) {
1642 1643
        /* Kill the AOF saving child as the AOF we already have may be longer
         * but contains the full dataset anyway. */
A
antirez 已提交
1644
        if (server.aof_child_pid != -1) {
1645 1646
            redisLog(REDIS_WARNING,
                "There is a child rewriting the AOF. Killing it!");
A
antirez 已提交
1647
            kill(server.aof_child_pid,SIGKILL);
1648
        }
1649
        /* Append only file: fsync() the AOF and exit */
1650
        redisLog(REDIS_NOTICE,"Calling fsync() on the AOF file.");
A
antirez 已提交
1651
        aof_fsync(server.aof_fd);
1652
    }
1653
    if ((server.saveparamslen > 0 && !nosave) || save) {
1654
        redisLog(REDIS_NOTICE,"Saving the final RDB snapshot before exiting.");
1655
        /* Snapshotting. Perform a SYNC SAVE and exit */
A
antirez 已提交
1656
        if (rdbSave(server.rdb_filename) != REDIS_OK) {
1657 1658 1659 1660 1661
            /* Ooops.. error saving! The best we can do is to continue
             * operating. Note that if there was a background saving process,
             * in the next cron() Redis will be notified that the background
             * saving aborted, handling special stuff like slaves pending for
             * synchronization... */
1662
            redisLog(REDIS_WARNING,"Error trying to save the DB, can't exit.");
1663 1664 1665
            return REDIS_ERR;
        }
    }
1666 1667 1668 1669
    if (server.daemonize) {
        redisLog(REDIS_NOTICE,"Removing the pid file.");
        unlink(server.pidfile);
    }
1670 1671 1672
    /* Close the listening sockets. Apparently this allows faster restarts. */
    if (server.ipfd != -1) close(server.ipfd);
    if (server.sofd != -1) close(server.sofd);
1673 1674 1675 1676
    if (server.unixsocket) {
        redisLog(REDIS_NOTICE,"Removing the unix socket file.");
        unlink(server.unixsocket); /* don't care if this fails */
    }
1677

1678
    redisLog(REDIS_WARNING,"Redis is now ready to exit, bye bye...");
1679 1680 1681 1682 1683 1684
    return REDIS_OK;
}

/*================================== Commands =============================== */

void authCommand(redisClient *c) {
1685 1686 1687
    if (!server.requirepass) {
        addReplyError(c,"Client sent AUTH, but no password is set");
    } else if (!strcmp(c->argv[1]->ptr, server.requirepass)) {
1688 1689 1690 1691
      c->authenticated = 1;
      addReply(c,shared.ok);
    } else {
      c->authenticated = 0;
1692
      addReplyError(c,"invalid password");
1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703
    }
}

void pingCommand(redisClient *c) {
    addReply(c,shared.pong);
}

void echoCommand(redisClient *c) {
    addReplyBulk(c,c->argv[1]);
}

A
antirez 已提交
1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714
void timeCommand(redisClient *c) {
    struct timeval tv;

    /* gettimeofday() can only fail if &tv is a bad addresss so we
     * don't check for errors. */
    gettimeofday(&tv,NULL);
    addReplyMultiBulkLen(c,2);
    addReplyBulkLongLong(c,tv.tv_sec);
    addReplyBulkLongLong(c,tv.tv_usec);
}

1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738
/* Convert an amount of bytes into a human readable string in the form
 * of 100B, 2G, 100M, 4K, and so forth. */
void bytesToHuman(char *s, unsigned long long n) {
    double d;

    if (n < 1024) {
        /* Bytes */
        sprintf(s,"%lluB",n);
        return;
    } else if (n < (1024*1024)) {
        d = (double)n/(1024);
        sprintf(s,"%.2fK",d);
    } else if (n < (1024LL*1024*1024)) {
        d = (double)n/(1024*1024);
        sprintf(s,"%.2fM",d);
    } else if (n < (1024LL*1024*1024*1024)) {
        d = (double)n/(1024LL*1024*1024);
        sprintf(s,"%.2fG",d);
    }
}

/* Create the string returned by the INFO command. This is decoupled
 * by the INFO command itself as we need to report the same information
 * on memory corruption problems. */
1739 1740
sds genRedisInfoString(char *section) {
    sds info = sdsempty();
1741
    time_t uptime = server.unixtime-server.stat_starttime;
A
antirez 已提交
1742
    int j, numcommands;
1743
    struct rusage self_ru, c_ru;
1744
    unsigned long lol, bib;
1745 1746 1747 1748 1749
    int allsections = 0, defsections = 0;
    int sections = 0;
    
    if (section) {
        allsections = strcasecmp(section,"all") == 0;
1750
        defsections = strcasecmp(section,"default") == 0;
1751
    }
1752 1753 1754

    getrusage(RUSAGE_SELF, &self_ru);
    getrusage(RUSAGE_CHILDREN, &c_ru);
1755
    getClientsMaxBuffers(&lol,&bib);
1756 1757 1758

    /* Server */
    if (allsections || defsections || !strcasecmp(section,"server")) {
1759 1760
        struct utsname name;

1761
        if (sections++) info = sdscat(info,"\r\n");
1762
        uname(&name);
1763
        info = sdscatprintf(info,
1764 1765 1766 1767
            "# Server\r\n"
            "redis_version:%s\r\n"
            "redis_git_sha1:%s\r\n"
            "redis_git_dirty:%d\r\n"
1768
            "os:%s %s %s\r\n"
1769
            "arch_bits:%d\r\n"
1770
            "multiplexing_api:%s\r\n"
A
antirez 已提交
1771
            "gcc_version:%d.%d.%d\r\n"
1772
            "process_id:%ld\r\n"
A
antirez 已提交
1773
            "run_id:%s\r\n"
1774 1775 1776 1777 1778 1779 1780
            "tcp_port:%d\r\n"
            "uptime_in_seconds:%ld\r\n"
            "uptime_in_days:%ld\r\n"
            "lru_clock:%ld\r\n",
            REDIS_VERSION,
            redisGitSHA1(),
            strtol(redisGitDirty(),NULL,10) > 0,
1781
            name.sysname, name.release, name.machine,
1782
            server.arch_bits,
1783
            aeGetApiName(),
A
antirez 已提交
1784 1785 1786 1787 1788
#ifdef __GNUC__
            __GNUC__,__GNUC_MINOR__,__GNUC_PATCHLEVEL__,
#else
            0,0,0,
#endif
1789
            (long) getpid(),
A
antirez 已提交
1790
            server.runid,
1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801
            server.port,
            uptime,
            uptime/(3600*24),
            (unsigned long) server.lruclock);
    }

    /* Clients */
    if (allsections || defsections || !strcasecmp(section,"clients")) {
        if (sections++) info = sdscat(info,"\r\n");
        info = sdscatprintf(info,
            "# Clients\r\n"
1802
            "connected_clients:%lu\r\n"
1803 1804 1805 1806 1807 1808 1809 1810 1811 1812
            "client_longest_output_list:%lu\r\n"
            "client_biggest_input_buf:%lu\r\n"
            "blocked_clients:%d\r\n",
            listLength(server.clients)-listLength(server.slaves),
            lol, bib,
            server.bpop_blocked_clients);
    }

    /* Memory */
    if (allsections || defsections || !strcasecmp(section,"memory")) {
1813 1814 1815 1816 1817
        char hmem[64];
        char peak_hmem[64];

        bytesToHuman(hmem,zmalloc_used_memory());
        bytesToHuman(peak_hmem,server.stat_peak_memory);
1818 1819 1820 1821 1822 1823
        if (sections++) info = sdscat(info,"\r\n");
        info = sdscatprintf(info,
            "# Memory\r\n"
            "used_memory:%zu\r\n"
            "used_memory_human:%s\r\n"
            "used_memory_rss:%zu\r\n"
1824 1825
            "used_memory_peak:%zu\r\n"
            "used_memory_peak_human:%s\r\n"
A
antirez 已提交
1826
            "used_memory_lua:%lld\r\n"
1827
            "mem_fragmentation_ratio:%.2f\r\n"
1828
            "mem_allocator:%s\r\n",
1829 1830 1831
            zmalloc_used_memory(),
            hmem,
            zmalloc_get_rss(),
1832 1833
            server.stat_peak_memory,
            peak_hmem,
A
antirez 已提交
1834
            ((long long)lua_gc(server.lua,LUA_GCCOUNT,0))*1024LL,
1835
            zmalloc_get_fragmentation_ratio(),
1836
            ZMALLOC_LIB
1837
            );
1838 1839
    }

1840 1841 1842
    /* Persistence */
    if (allsections || defsections || !strcasecmp(section,"persistence")) {
        if (sections++) info = sdscat(info,"\r\n");
1843
        info = sdscatprintf(info,
1844 1845 1846 1847 1848 1849
            "# Persistence\r\n"
            "loading:%d\r\n"
            "aof_enabled:%d\r\n"
            "changes_since_last_save:%lld\r\n"
            "bgsave_in_progress:%d\r\n"
            "last_save_time:%ld\r\n"
1850
            "last_bgsave_status:%s\r\n"
1851 1852
            "bgrewriteaof_in_progress:%d\r\n"
            "bgrewriteaof_scheduled:%d\r\n",
1853
            server.loading,
1854
            server.aof_state != REDIS_AOF_OFF,
1855
            server.dirty,
A
antirez 已提交
1856
            server.rdb_child_pid != -1,
1857
            server.lastsave,
1858
            server.lastbgsave_status == REDIS_OK ? "ok" : "err",
1859 1860
            server.aof_child_pid != -1,
            server.aof_rewrite_scheduled);
1861

1862
        if (server.aof_state != REDIS_AOF_OFF) {
1863 1864 1865
            info = sdscatprintf(info,
                "aof_current_size:%lld\r\n"
                "aof_base_size:%lld\r\n"
1866 1867
                "aof_pending_rewrite:%d\r\n"
                "aof_buffer_length:%zu\r\n"
1868 1869
                "aof_pending_bio_fsync:%llu\r\n"
                "aof_delayed_fsync:%lu\r\n",
1870 1871 1872
                (long long) server.aof_current_size,
                (long long) server.aof_rewrite_base_size,
                server.aof_rewrite_scheduled,
A
antirez 已提交
1873
                sdslen(server.aof_buf),
1874 1875
                bioPendingJobsOfType(REDIS_BIO_AOF_FSYNC),
                server.aof_delayed_fsync);
1876 1877
        }

1878 1879 1880 1881 1882 1883 1884 1885 1886
        if (server.loading) {
            double perc;
            time_t eta, elapsed;
            off_t remaining_bytes = server.loading_total_bytes-
                                    server.loading_loaded_bytes;

            perc = ((double)server.loading_loaded_bytes /
                   server.loading_total_bytes) * 100;

1887
            elapsed = server.unixtime-server.loading_start_time;
1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904 1905 1906 1907
            if (elapsed == 0) {
                eta = 1; /* A fake 1 second figure if we don't have
                            enough info */
            } else {
                eta = (elapsed*remaining_bytes)/server.loading_loaded_bytes;
            }

            info = sdscatprintf(info,
                "loading_start_time:%ld\r\n"
                "loading_total_bytes:%llu\r\n"
                "loading_loaded_bytes:%llu\r\n"
                "loading_loaded_perc:%.2f\r\n"
                "loading_eta_seconds:%ld\r\n"
                ,(unsigned long) server.loading_start_time,
                (unsigned long long) server.loading_total_bytes,
                (unsigned long long) server.loading_loaded_bytes,
                perc,
                eta
            );
        }
1908
    }
1909 1910 1911 1912

    /* Stats */
    if (allsections || defsections || !strcasecmp(section,"stats")) {
        if (sections++) info = sdscat(info,"\r\n");
1913
        info = sdscatprintf(info,
1914 1915 1916
            "# Stats\r\n"
            "total_connections_received:%lld\r\n"
            "total_commands_processed:%lld\r\n"
1917
            "instantaneous_ops_per_sec:%lld\r\n"
1918
            "rejected_connections:%lld\r\n"
1919 1920 1921 1922 1923
            "expired_keys:%lld\r\n"
            "evicted_keys:%lld\r\n"
            "keyspace_hits:%lld\r\n"
            "keyspace_misses:%lld\r\n"
            "pubsub_channels:%ld\r\n"
1924
            "pubsub_patterns:%lu\r\n"
1925
            "latest_fork_usec:%lld\r\n",
1926 1927
            server.stat_numconnections,
            server.stat_numcommands,
1928
            getOperationsPerSecond(),
1929
            server.stat_rejected_conn,
1930 1931 1932 1933 1934
            server.stat_expiredkeys,
            server.stat_evictedkeys,
            server.stat_keyspace_hits,
            server.stat_keyspace_misses,
            dictSize(server.pubsub_channels),
1935 1936
            listLength(server.pubsub_patterns),
            server.stat_fork_time);
1937
    }
A
antirez 已提交
1938

1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954
    /* Replication */
    if (allsections || defsections || !strcasecmp(section,"replication")) {
        if (sections++) info = sdscat(info,"\r\n");
        info = sdscatprintf(info,
            "# Replication\r\n"
            "role:%s\r\n",
            server.masterhost == NULL ? "master" : "slave");
        if (server.masterhost) {
            info = sdscatprintf(info,
                "master_host:%s\r\n"
                "master_port:%d\r\n"
                "master_link_status:%s\r\n"
                "master_last_io_seconds_ago:%d\r\n"
                "master_sync_in_progress:%d\r\n"
                ,server.masterhost,
                server.masterport,
A
antirez 已提交
1955
                (server.repl_state == REDIS_REPL_CONNECTED) ?
1956 1957
                    "up" : "down",
                server.master ?
1958
                ((int)(server.unixtime-server.master->lastinteraction)) : -1,
A
antirez 已提交
1959
                server.repl_state == REDIS_REPL_TRANSFER
1960 1961
            );

A
antirez 已提交
1962
            if (server.repl_state == REDIS_REPL_TRANSFER) {
1963 1964 1965 1966
                info = sdscatprintf(info,
                    "master_sync_left_bytes:%ld\r\n"
                    "master_sync_last_io_seconds_ago:%d\r\n"
                    ,(long)server.repl_transfer_left,
1967
                    (int)(server.unixtime-server.repl_transfer_lastio)
1968 1969
                );
            }
1970

A
antirez 已提交
1971
            if (server.repl_state != REDIS_REPL_CONNECTED) {
1972 1973
                info = sdscatprintf(info,
                    "master_link_down_since_seconds:%ld\r\n",
1974
                    (long)server.unixtime-server.repl_down_since);
1975
            }
A
antirez 已提交
1976
        }
1977
        info = sdscatprintf(info,
1978
            "connected_slaves:%lu\r\n",
1979
            listLength(server.slaves));
1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
        if (listLength(server.slaves)) {
            int slaveid = 0;
            listNode *ln;
            listIter li;

            listRewind(server.slaves,&li);
            while((ln = listNext(&li))) {
                redisClient *slave = listNodeValue(ln);
                char *state = NULL;
                char ip[32];
                int port;

                if (anetPeerToString(slave->fd,ip,&port) == -1) continue;
                switch(slave->replstate) {
                case REDIS_REPL_WAIT_BGSAVE_START:
                case REDIS_REPL_WAIT_BGSAVE_END:
                    state = "wait_bgsave";
                    break;
                case REDIS_REPL_SEND_BULK:
                    state = "send_bulk";
                    break;
                case REDIS_REPL_ONLINE:
                    state = "online";
                    break;
                }
                if (state == NULL) continue;
                info = sdscatprintf(info,"slave%d:%s,%d,%s\r\n",
                    slaveid,ip,port,state);
                slaveid++;
            }
        }
A
antirez 已提交
2011 2012
    }

2013 2014
    /* CPU */
    if (allsections || defsections || !strcasecmp(section,"cpu")) {
2015 2016
        if (sections++) info = sdscat(info,"\r\n");
        info = sdscatprintf(info,
2017
        "# CPU\r\n"
2018 2019
        "used_cpu_sys:%.2f\r\n"
        "used_cpu_user:%.2f\r\n"
2020 2021
        "used_cpu_sys_children:%.2f\r\n"
        "used_cpu_user_children:%.2f\r\n",
2022
        (float)self_ru.ru_stime.tv_sec+(float)self_ru.ru_stime.tv_usec/1000000,
2023 2024 2025
        (float)self_ru.ru_utime.tv_sec+(float)self_ru.ru_utime.tv_usec/1000000,
        (float)c_ru.ru_stime.tv_sec+(float)c_ru.ru_stime.tv_usec/1000000,
        (float)c_ru.ru_utime.tv_sec+(float)c_ru.ru_utime.tv_usec/1000000);
2026
    }
2027

2028 2029 2030 2031
    /* cmdtime */
    if (allsections || !strcasecmp(section,"commandstats")) {
        if (sections++) info = sdscat(info,"\r\n");
        info = sdscatprintf(info, "# Commandstats\r\n");
2032
        numcommands = sizeof(redisCommandTable)/sizeof(struct redisCommand);
2033
        for (j = 0; j < numcommands; j++) {
2034
            struct redisCommand *c = redisCommandTable+j;
2035

2036 2037 2038 2039 2040
            if (!c->calls) continue;
            info = sdscatprintf(info,
                "cmdstat_%s:calls=%lld,usec=%lld,usec_per_call=%.2f\r\n",
                c->name, c->calls, c->microseconds,
                (c->calls == 0) ? 0 : ((float)c->microseconds/c->calls));
2041
        }
A
antirez 已提交
2042 2043
    }

2044
    /* Cluster */
2045 2046 2047 2048 2049 2050 2051 2052
    if (allsections || defsections || !strcasecmp(section,"cluster")) {
        if (sections++) info = sdscat(info,"\r\n");
        info = sdscatprintf(info,
        "# Cluster\r\n"
        "cluster_enabled:%d\r\n",
        server.cluster_enabled);
    }

2053 2054 2055 2056 2057 2058
    /* Key space */
    if (allsections || defsections || !strcasecmp(section,"keyspace")) {
        if (sections++) info = sdscat(info,"\r\n");
        info = sdscatprintf(info, "# Keyspace\r\n");
        for (j = 0; j < server.dbnum; j++) {
            long long keys, vkeys;
2059

2060 2061 2062 2063 2064 2065
            keys = dictSize(server.db[j].dict);
            vkeys = dictSize(server.db[j].expires);
            if (keys || vkeys) {
                info = sdscatprintf(info, "db%d:keys=%lld,expires=%lld\r\n",
                    j, keys, vkeys);
            }
2066 2067 2068 2069 2070 2071
        }
    }
    return info;
}

void infoCommand(redisClient *c) {
2072 2073 2074 2075 2076 2077 2078
    char *section = c->argc == 2 ? c->argv[1]->ptr : "default";

    if (c->argc > 2) {
        addReply(c,shared.syntaxerr);
        return;
    }
    sds info = genRedisInfoString(section);
2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097
    addReplySds(c,sdscatprintf(sdsempty(),"$%lu\r\n",
        (unsigned long)sdslen(info)));
    addReplySds(c,info);
    addReply(c,shared.crlf);
}

void monitorCommand(redisClient *c) {
    /* ignore MONITOR if aleady slave or in monitor mode */
    if (c->flags & REDIS_SLAVE) return;

    c->flags |= (REDIS_SLAVE|REDIS_MONITOR);
    c->slaveseldb = 0;
    listAddNodeTail(server.monitors,c);
    addReply(c,shared.ok);
}

/* ============================ Maxmemory directive  ======================== */

/* This function gets called when 'maxmemory' is set on the config file to limit
2098
 * the max memory used by the server, before processing a command.
2099
 *
2100 2101
 * The goal of the function is to free enough memory to keep Redis under the
 * configured memory limit.
2102
 *
2103 2104 2105 2106 2107 2108 2109 2110
 * The function starts calculating how many bytes should be freed to keep
 * Redis under the limit, and enters a loop selecting the best keys to
 * evict accordingly to the configured policy.
 *
 * If all the bytes needed to return back under the limit were freed the
 * function returns REDIS_OK, otherwise REDIS_ERR is returned, and the caller
 * should block the execution of commands that will result in more memory
 * used by the server.
2111
 */
2112 2113 2114 2115
int freeMemoryIfNeeded(void) {
    size_t mem_used, mem_tofree, mem_freed;
    int slaves = listLength(server.slaves);

2116 2117
    /* Remove the size of slaves output buffers and AOF buffer from the
     * count of used memory. */
2118 2119 2120 2121 2122 2123 2124 2125 2126 2127 2128 2129 2130 2131 2132
    mem_used = zmalloc_used_memory();
    if (slaves) {
        listIter li;
        listNode *ln;

        listRewind(server.slaves,&li);
        while((ln = listNext(&li))) {
            redisClient *slave = listNodeValue(ln);
            unsigned long obuf_bytes = getClientOutputBufferMemoryUsage(slave);
            if (obuf_bytes > mem_used)
                mem_used = 0;
            else
                mem_used -= obuf_bytes;
        }
    }
2133 2134 2135 2136
    if (server.aof_state != REDIS_AOF_OFF) {
        mem_used -= sdslen(server.aof_buf);
        mem_used -= sdslen(server.aof_rewrite_buf);
    }
2137

2138 2139 2140 2141 2142 2143 2144 2145 2146 2147 2148
    /* Check if we are over the memory limit. */
    if (mem_used <= server.maxmemory) return REDIS_OK;

    if (server.maxmemory_policy == REDIS_MAXMEMORY_NO_EVICTION)
        return REDIS_ERR; /* We need to free memory, but policy forbids. */

    /* Compute how much memory we need to free. */
    mem_tofree = mem_used - server.maxmemory;
    mem_freed = 0;
    while (mem_freed < mem_tofree) {
        int j, k, keys_freed = 0;
2149

2150
        for (j = 0; j < server.dbnum; j++) {
2151
            long bestval = 0; /* just to prevent warning */
2152 2153 2154 2155 2156 2157 2158 2159 2160 2161 2162 2163 2164 2165 2166 2167 2168 2169 2170
            sds bestkey = NULL;
            struct dictEntry *de;
            redisDb *db = server.db+j;
            dict *dict;

            if (server.maxmemory_policy == REDIS_MAXMEMORY_ALLKEYS_LRU ||
                server.maxmemory_policy == REDIS_MAXMEMORY_ALLKEYS_RANDOM)
            {
                dict = server.db[j].dict;
            } else {
                dict = server.db[j].expires;
            }
            if (dictSize(dict) == 0) continue;

            /* volatile-random and allkeys-random policy */
            if (server.maxmemory_policy == REDIS_MAXMEMORY_ALLKEYS_RANDOM ||
                server.maxmemory_policy == REDIS_MAXMEMORY_VOLATILE_RANDOM)
            {
                de = dictGetRandomKey(dict);
2171
                bestkey = dictGetKey(de);
2172 2173 2174 2175 2176 2177 2178 2179 2180 2181 2182 2183
            }

            /* volatile-lru and allkeys-lru policy */
            else if (server.maxmemory_policy == REDIS_MAXMEMORY_ALLKEYS_LRU ||
                server.maxmemory_policy == REDIS_MAXMEMORY_VOLATILE_LRU)
            {
                for (k = 0; k < server.maxmemory_samples; k++) {
                    sds thiskey;
                    long thisval;
                    robj *o;

                    de = dictGetRandomKey(dict);
2184
                    thiskey = dictGetKey(de);
2185 2186 2187 2188
                    /* When policy is volatile-lru we need an additonal lookup
                     * to locate the real key, as dict is set to db->expires. */
                    if (server.maxmemory_policy == REDIS_MAXMEMORY_VOLATILE_LRU)
                        de = dictFind(db->dict, thiskey);
2189
                    o = dictGetVal(de);
2190 2191 2192 2193 2194 2195 2196 2197 2198 2199 2200 2201 2202 2203 2204 2205 2206
                    thisval = estimateObjectIdleTime(o);

                    /* Higher idle time is better candidate for deletion */
                    if (bestkey == NULL || thisval > bestval) {
                        bestkey = thiskey;
                        bestval = thisval;
                    }
                }
            }

            /* volatile-ttl */
            else if (server.maxmemory_policy == REDIS_MAXMEMORY_VOLATILE_TTL) {
                for (k = 0; k < server.maxmemory_samples; k++) {
                    sds thiskey;
                    long thisval;

                    de = dictGetRandomKey(dict);
2207 2208
                    thiskey = dictGetKey(de);
                    thisval = (long) dictGetVal(de);
2209 2210 2211 2212 2213 2214 2215 2216 2217 2218 2219 2220

                    /* Expire sooner (minor expire unix timestamp) is better
                     * candidate for deletion */
                    if (bestkey == NULL || thisval < bestval) {
                        bestkey = thiskey;
                        bestval = thisval;
                    }
                }
            }

            /* Finally remove the selected key. */
            if (bestkey) {
2221 2222
                long long delta;

2223
                robj *keyobj = createStringObject(bestkey,sdslen(bestkey));
2224
                propagateExpire(db,keyobj);
2225 2226 2227 2228 2229 2230 2231 2232 2233
                /* We compute the amount of memory freed by dbDelete() alone.
                 * It is possible that actually the memory needed to propagate
                 * the DEL in AOF and replication link is greater than the one
                 * we are freeing removing the key, but we can't account for
                 * that otherwise we would never exit the loop.
                 *
                 * AOF and Output buffer memory will be freed eventually so
                 * we only care about memory used by the key space. */
                delta = (long long) zmalloc_used_memory();
2234
                dbDelete(db,keyobj);
2235 2236
                delta -= (long long) zmalloc_used_memory();
                mem_freed += delta;
2237
                server.stat_evictedkeys++;
2238
                decrRefCount(keyobj);
2239 2240 2241 2242 2243 2244
                keys_freed++;

                /* When the memory to free starts to be big enough, we may
                 * start spending so much time here that is impossible to
                 * deliver data to the slaves fast enough, so we force the
                 * transmission here inside the loop. */
A
antirez 已提交
2245
                if (slaves) flushSlavesOutputBuffers();
2246 2247
            }
        }
2248
        if (!keys_freed) return REDIS_ERR; /* nothing to free... */
2249
    }
2250
    return REDIS_OK;
2251 2252 2253 2254 2255 2256 2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 2267 2268 2269 2270 2271 2272 2273 2274 2275 2276
}

/* =================================== Main! ================================ */

#ifdef __linux__
int linuxOvercommitMemoryValue(void) {
    FILE *fp = fopen("/proc/sys/vm/overcommit_memory","r");
    char buf[64];

    if (!fp) return -1;
    if (fgets(buf,64,fp) == NULL) {
        fclose(fp);
        return -1;
    }
    fclose(fp);

    return atoi(buf);
}

void linuxOvercommitMemoryWarning(void) {
    if (linuxOvercommitMemoryValue() == 0) {
        redisLog(REDIS_WARNING,"WARNING overcommit_memory is set to 0! Background save may fail under low memory condition. To fix this issue add 'vm.overcommit_memory = 1' to /etc/sysctl.conf and then reboot or run the command 'sysctl vm.overcommit_memory=1' for this to take effect.");
    }
}
#endif /* __linux__ */

2277 2278 2279 2280
void createPidFile(void) {
    /* Try to write the pid file in a best-effort way. */
    FILE *fp = fopen(server.pidfile,"w");
    if (fp) {
2281
        fprintf(fp,"%d\n",(int)getpid());
2282 2283 2284 2285
        fclose(fp);
    }
}

2286 2287 2288 2289 2290 2291 2292 2293 2294 2295 2296 2297 2298 2299 2300 2301 2302 2303
void daemonize(void) {
    int fd;

    if (fork() != 0) exit(0); /* parent exits */
    setsid(); /* create a new session */

    /* Every output goes to /dev/null. If Redis is daemonized but
     * the 'logfile' is set to 'stdout' in the configuration file
     * it will not log at all. */
    if ((fd = open("/dev/null", O_RDWR, 0)) != -1) {
        dup2(fd, STDIN_FILENO);
        dup2(fd, STDOUT_FILENO);
        dup2(fd, STDERR_FILENO);
        if (fd > STDERR_FILENO) close(fd);
    }
}

void version() {
A
antirez 已提交
2304 2305 2306 2307 2308 2309
    printf("Redis server v=%s sha=%s:%d malloc=%s bits=%d\n",
        REDIS_VERSION,
        redisGitSHA1(),
        atoi(redisGitDirty()) > 0,
        ZMALLOC_LIB,
        sizeof(long) == 4 ? 32 : 64);
2310 2311 2312 2313
    exit(0);
}

void usage() {
2314
    fprintf(stderr,"Usage: ./redis-server [/path/to/redis.conf] [options]\n");
2315
    fprintf(stderr,"       ./redis-server - (read config from stdin)\n");
2316
    fprintf(stderr,"       ./redis-server -v or --version\n");
2317 2318
    fprintf(stderr,"       ./redis-server -h or --help\n");
    fprintf(stderr,"       ./redis-server --test-memory <megabytes>\n\n");
2319 2320 2321 2322 2323 2324
    fprintf(stderr,"Examples:\n");
    fprintf(stderr,"       ./redis-server (run the server with default conf)\n");
    fprintf(stderr,"       ./redis-server /etc/redis/6379.conf\n");
    fprintf(stderr,"       ./redis-server --port 7777\n");
    fprintf(stderr,"       ./redis-server --port 7777 --slaveof 127.0.0.1 8888\n");
    fprintf(stderr,"       ./redis-server /etc/myredis.conf --loglevel verbose\n");
2325 2326 2327
    exit(1);
}

A
antirez 已提交
2328 2329 2330 2331 2332 2333 2334 2335 2336 2337 2338 2339 2340 2341 2342 2343 2344
void redisAsciiArt(void) {
#include "asciilogo.h"
    char *buf = zmalloc(1024*16);

    snprintf(buf,1024*16,ascii_logo,
        REDIS_VERSION,
        redisGitSHA1(),
        strtol(redisGitDirty(),NULL,10) > 0,
        (sizeof(long) == 8) ? "64" : "32",
        server.cluster_enabled ? "cluster" : "stand alone",
        server.port,
        (long) getpid()
    );
    redisLogRaw(REDIS_NOTICE|REDIS_LOG_RAW,buf);
    zfree(buf);
}

2345
static void sigtermHandler(int sig) {
2346 2347
    REDIS_NOTUSED(sig);

A
antirez 已提交
2348
    redisLogFromHandler(REDIS_WARNING,"Received SIGTERM, scheduling shutdown...");
2349 2350 2351
    server.shutdown_asap = 1;
}

2352
void setupSignalHandlers(void) {
2353
    struct sigaction act;
2354 2355 2356 2357 2358 2359 2360
    stack_t stack;

    stack.ss_sp = altstack;
    stack.ss_flags = 0;
    stack.ss_size = SIGSTKSZ;

    sigaltstack(&stack, NULL);
2361

2362 2363 2364
    /* When the SA_SIGINFO flag is set in sa_flags then sa_sigaction is used.
     * Otherwise, sa_handler is used. */
    sigemptyset(&act.sa_mask);
2365
    act.sa_flags = 0;
2366
    act.sa_handler = sigtermHandler;
2367
    sigaction(SIGTERM, &act, NULL);
2368

2369
#ifdef HAVE_BACKTRACE
2370 2371
    /* Use alternate stack so we don't clobber stack in case of segv, or when we run out of stack ..
     * also resethand & nodefer so we can get interrupted (and killed) if we cause SEGV during SEGV handler */
2372 2373 2374 2375 2376 2377 2378 2379 2380
    sigemptyset(&act.sa_mask);
    act.sa_flags = SA_NODEFER | SA_ONSTACK | SA_RESETHAND | SA_SIGINFO;
    act.sa_sigaction = sigsegvHandler;
    sigaction(SIGSEGV, &act, NULL);
    sigaction(SIGBUS, &act, NULL);
    sigaction(SIGFPE, &act, NULL);
    sigaction(SIGILL, &act, NULL);
#endif
    return;
2381 2382
}

2383 2384
void memtest(size_t megabytes, int passes);

2385 2386
int main(int argc, char **argv) {
    long long start;
2387
    struct timeval tv;
2388

A
antirez 已提交
2389
    /* We need to initialize our libraries, and the server configuration. */
2390
    zmalloc_enable_thread_safeness();
2391 2392 2393
    srand(time(NULL)^getpid());
    gettimeofday(&tv,NULL);
    dictSetHashFunctionSeed(tv.tv_sec^tv.tv_usec^getpid());
2394
    initServerConfig();
2395

2396 2397 2398 2399 2400 2401
    if (argc >= 2) {
        int j = 1; /* First option to parse in argv[] */
        sds options = sdsempty();
        char *configfile = NULL;

        /* Handle special options --help and --version */
2402 2403
        if (strcmp(argv[1], "-v") == 0 ||
            strcmp(argv[1], "--version") == 0) version();
2404 2405
        if (strcmp(argv[1], "--help") == 0 ||
            strcmp(argv[1], "-h") == 0) usage();
2406 2407
        if (strcmp(argv[1], "--test-memory") == 0) {
            if (argc == 3) {
2408
                memtest(atoi(argv[2]),50);
2409 2410 2411 2412 2413 2414 2415 2416
                exit(0);
            } else {
                fprintf(stderr,"Please specify the amount of memory to test in megabytes.\n");
                fprintf(stderr,"Example: ./redis-server --test-memory 4096\n\n");
                exit(1);
            }
        }

2417 2418 2419 2420 2421 2422 2423 2424 2425 2426 2427 2428 2429 2430 2431 2432 2433 2434 2435 2436
        /* First argument is the config file name? */
        if (argv[j][0] != '-' || argv[j][1] != '-')
            configfile = argv[j++];
        /* All the other options are parsed and conceptually appended to the
         * configuration file. For instance --port 6380 will generate the
         * string "port 6380\n" to be parsed after the actual file name
         * is parsed, if any. */
        while(j != argc) {
            if (argv[j][0] == '-' && argv[j][1] == '-') {
                /* Option name */
                if (sdslen(options)) options = sdscat(options,"\n");
                options = sdscat(options,argv[j]+2);
                options = sdscat(options," ");
            } else {
                /* Option argument */
                options = sdscatrepr(options,argv[j],strlen(argv[j]));
                options = sdscat(options," ");
            }
            j++;
        }
2437
        resetServerSaveParams();
2438 2439
        loadServerConfig(configfile,options);
        sdsfree(options);
2440 2441 2442 2443 2444 2445 2446
    } else {
        redisLog(REDIS_WARNING,"Warning: no config file specified, using the default config. In order to specify a config file use 'redis-server /path/to/redis.conf'");
    }
    if (server.daemonize) daemonize();
    initServer();
    if (server.daemonize) createPidFile();
    redisAsciiArt();
2447
    redisLog(REDIS_WARNING,"Server started, Redis version " REDIS_VERSION);
2448 2449 2450 2451
#ifdef __linux__
    linuxOvercommitMemoryWarning();
#endif
    start = ustime();
2452
    if (server.aof_state == REDIS_AOF_ON) {
2453
        if (loadAppendOnlyFile(server.aof_filename) == REDIS_OK)
2454 2455
            redisLog(REDIS_NOTICE,"DB loaded from append only file: %.3f seconds",(float)(ustime()-start)/1000000);
    } else {
A
antirez 已提交
2456
        if (rdbLoad(server.rdb_filename) == REDIS_OK) {
2457 2458 2459 2460 2461 2462 2463 2464 2465 2466 2467 2468 2469 2470 2471 2472 2473
            redisLog(REDIS_NOTICE,"DB loaded from disk: %.3f seconds",
                (float)(ustime()-start)/1000000);
        } else if (errno != ENOENT) {
            redisLog(REDIS_WARNING,"Fatal error loading the DB. Exiting.");
            exit(1);
        }
    }
    if (server.ipfd > 0)
        redisLog(REDIS_NOTICE,"The server is now ready to accept connections on port %d", server.port);
    if (server.sofd > 0)
        redisLog(REDIS_NOTICE,"The server is now ready to accept connections at %s", server.unixsocket);
    aeSetBeforeSleepProc(server.el,beforeSleep);
    aeMain(server.el);
    aeDeleteEventLoop(server.el);
    return 0;
}

2474
/* The End */